Content area
Due to differences in physiological characteristics and drug metabolism between children and adults, drug efficacy evaluation and safety monitoring in pediatric drug development present significant challenges. This paper proposes a data-driven incentive mechanism for pediatric drug development based on medical imaging data. This approach optimizes drug market pricing through precise imaging data, promoting accessibility and R&D efficiency for pediatric drugs. This study first collects multi-source computed tomography (CT), magnetic resonance imaging (MRI), and X-ray data, focusing on images of common pediatric diseases. After data preprocessing, a convolutional neural network (CNN) is used for feature extraction to extract key image information. Image difference methods and a U-Net image segmentation network are then used to evaluate drug efficacy and safety, quantify efficacy changes, and analyze side effects. Next, a drug efficacy-safety evaluation model is developed, and game theory is employed to design a R&D incentive mechanism. Monte Carlo simulation is combined with risk assessment to comprehensively consider factors such as cost, R&D investment, and market demand during the pricing optimization phase. A dynamic pricing strategy is implemented to ensure both economic benefits and social accessibility of the drug. Experiments have shown that the drug has a good development effect, with an average tumor volume reduction of 32.7% (95% CI: 28.4%-36.9%). The drug’s impact on organ volume is within ± 2 cm³, and the market pricing strategy selects a relatively optimal price point.
Introduction
With the increase in the number of children’s diseases and the complexity of drug research and development (R&D) worldwide, the R&D of children’s drugs faces more and more challenges1,2. The existing drug R&D and evaluation system is usually based on the clinical data of adult patients and lacks special design and optimization for children. Individual differences may exist in the response of children’s patients to medications3. Therefore, relying only on data from studies of adult drugs makes it difficult to effectively assess the suitability of drugs for the children’s population4,5. Furthermore, the research and development of children’s drugs usually requires more time and resources, and the high risk makes pharmaceutical companies’ investment in this area more limited. Due to the relatively small market size of children’s drugs, pharmaceutical companies usually lack sufficient motivation for children’s drug R&D, resulting in a lag in the research and development of children’s drugs, and the price of drugs is usually high, further affecting the accessibility and popularization of children’s drugs6, 7–8. Therefore, effective strategies are urgently needed to improve the efficiency of research and development of children’s drugs, minimize risks, and ensure that drugs are priced appropriately in the market.
With the rapid development of medical imaging technology, its application in drug research and development has received increasing attention. Medical imaging technology, especially CT (Computed Tomography), MRI (Magnetic Resonance Imaging), X-ray, and other imaging methods, provides precise information on the structure and pathological conditions of in vivo tissues, and therefore plays an important role in the evaluation of drug efficacy9,10. Many researchers have tried to use these technologies to detect the therapeutic effects of drugs on children’s diseases (such as cancer, bone diseases, and nervous system diseases). Through imaging data, the changes in lesions before and after drug treatment can be precisely measured to quantify the efficacy of drugs. Moreover, medical imaging is also used to evaluate the safety of drugs and monitor whether drugs cause side effects on other organs11,12. For example, CT and MRI images help analyze the potential damage of drugs to organs and detect and evaluate side effects in advance. However, most existing imaging data analysis methods are limited to basic visual assessments and lack intelligent and systematic analysis tools. In particular, imaging data analysis methods for children are still relatively scarce13, 14–15 and are in urgent need of further optimization and development. In particular, in the R&D of children’s drugs, how to use precise imaging data for efficacy evaluation and safety monitoring, and how to combine these data with the incentive mechanism for drug R&D are still important issues that need to be addressed.
To address the above problems, some researchers have tried to combine machine learning and artificial intelligence methods with medical imaging technology to improve the evaluation of drug efficacy and safety. CNN is widely used in medical image processing. By extracting and analyzing features from image data, it helps people automatically identify the effects of drug treatment, especially in the detection of target areas such as tumors and lesions16. The research of Xu Y et al.17 emphasized the application of medical databases and machine learning (including supervised learning and unsupervised learning) in clinical big data for drug compliance. In addition, deep learning models such as U-Net have performed well in medical image segmentation, precisely delineating lesion areas and improving the evaluation accuracy. With the accumulation of a large amount of medical image data and the continuous advancement of deep learning technology, the automated processing and analysis of image data has become an indispensable tool in drug R&D. However, although the existing image data processing methods improve diagnostic accuracy, in terms of the design of drug R&D incentive mechanisms, most existing studies only focus on quantitative evaluation of drug efficacy and safety, ignoring the practical application of how to motivate pharmaceutical companies to increase R&D investment through data-driven18,19. At the same time, most of the existing pricing optimization methods do not fully consider the dynamic changes of drugs in the market, resulting in limited application effects of pricing models. To better motivate pharmaceutical companies to conduct R&D of children’s drugs, it is necessary to combine medical imaging technology with the economic benefits of drug R&D, so as to develop a comprehensive incentive mechanism that truly promotes children’s drug R&D20, 21–22.
The purpose of this study is to propose a children’s drug R&D incentive mechanism driven by medical imaging data, and optimize drug market pricing through precise imaging data to promote the R&D efficiency and market accessibility of children’s drugs. Specifically, this paper collects different types of medical imaging data, performs precise preprocessing on them, uses CNN and U-Net networks for feature extraction and image segmentation, and evaluates the efficacy and safety of drugs. On this basis, this paper combines game theory to design a reasonable R&D incentive mechanism to encourage pharmaceutical companies to increase their R&D investment in children’s drugs. At the same time, based on image data analysis, this paper also builds a dynamic pricing model, combining market demand, cost, R&D investment, and other factors to optimize the pricing strategy of drugs to ensure that drugs are both competitive in the market and maintain a reasonable pricing level. Through these methods, this paper aims to solve the high-risk and low-return problems currently faced by children’s drug R&D, improve the efficiency of drug R&D, and promote the accessibility of children’s drugs in the market, bringing more sustainable solutions to the pharmaceutical industry, policymakers, and society. These research results can not only provide technical support for pharmaceutical companies, but also provide data-driven decision-making basis for policymakers to promote the long-term development of children’s drug R&D.
Methods
Data collection and preprocessing
This study collects medical imaging data from multiple hospitals and research institutions, including CT, MRI, X-ray, etc., focusing on common children’s diseases (such as cancer, bone diseases, neurological diseases, etc.). The data comes from different equipment and scanning conditions to ensure diversity and representativeness. All data comply with ethical review and privacy protection requirements to ensure legality and standardization. Data preprocessing aims to remove noise, ensure high-quality input, and improve analysis accuracy.
During the data collection process, the resolution of CT and MRI images is generally set to 1 mm or higher to ensure clear presentation of details of the diseased area. For X-ray images, the resolution is usually set to 0.1 mm to ensure that subtle changes in bones or other tissues are captured. The annotation of image data is completed by professionally trained medical imaging experts to ensure the accuracy and consistency of the annotation23,24. In addition, to establish a more comprehensive dataset, this study also collects imaging data of children of different ages, covering different groups from infants to adolescents, to take into account the impact of physiological differences on drug response.
The core task of data preprocessing is to perform a series of processing on the original image to eliminate noise, enhance image details, and unify data size and resolution to ensure that the data meets the requirements of the subsequent analysis model. The specific steps include the following aspects:
Noise removal and filtering: first, the Gaussian filter algorithm is used to denoise the image. The Gaussian filter effectively removes random noise caused by imaging equipment and environmental factors, thereby improving the quality of the image. The Gaussian kernel size used is generally 5 × 5 to 7 × 7 pixels, with a standard deviation of 1.0 to 1.5. The specific parameters are adjusted according to the image quality and noise level. The image adjusted by Gaussian filtering is shown in Fig. 1. The image after Gaussian filtering can provide a clearer image basis for subsequent image segmentation and analysis, thereby improving the accuracy of drug R&D and efficacy evaluation.
This study includes imaging data from 6 tertiary pediatric hospitals, covering 5,382 children (age stratification: 1,287 cases aged 0–2 years, 1,980 cases aged 3–6 years, 1,678 cases aged 7–12 years, 437 cases aged 13–18 years) and indications such as tumors, bone metabolic diseases, and epilepsy. The age-stratified analysis uses the ANCOVA model (with age as a covariate) to quantitatively verify the metabolic differences between infants and adolescents (p < 0.01). Patient images are divided into training/validation/test sets at 7:1.5:1.5, and age-stratified sampling ensures balance.
Fig. 1 [Images not available. See PDF.]
Noise reduction of CT and MRI images.
This study uses the Canny edge detection algorithm to extract key features in the image, such as tumors, lesions, and organ contours. The algorithm achieves precise edge detection through smoothing, gradient calculation, non-maximum suppression, and dual threshold detection. To cope with the differences in image size and resolution, normalization (scaling pixel values to [0, 1]) and standardization (adjusting brightness and contrast through mean and standard deviation) are performed to ensure consistency between different types of images (CT, MRI, X-ray).
Size and resolution adjustment: since the size and resolution of medical images are often inconsistent, all images need to be uniformly sized and adjusted. Usually, the adjusted image size is 256 × 256 pixels, and the resolution is set to 1 mm/pixel. For images with lower resolution, bilinear interpolation is used to increase the resolution to ensure precision in subsequent analysis. Through these preprocessing steps, the quality of image data is significantly improved, providing reliable data support for subsequent feature extraction and model training.
Feature extraction is a crucial step in medical image analysis. The purpose is to extract key information that is meaningful for drug efficacy evaluation and safety monitoring from preprocessed medical images. In this study, CNN is used for feature extraction. CNN is a deep learning algorithm that automatically extracts hierarchical features from input data and has been proven to be extremely accurate and robust in medical image processing.
Specifically, the convolutional layer of CNN effectively extracts local features in the image, such as tumor boundaries, organ contours, and lesion areas, through convolution operations with filters. To improve the accuracy of feature extraction, this study adopts the ResNet structure, which is a deep residual network that effectively solves the gradient vanishing problem in deep network training and improves training efficiency. The depth of the network is set to 50 layers to fully mine the complex features in the image data. After multiple convolution and pooling operations, the network finally outputs high-level feature representations of each image, which is used as input for subsequent analysis.
In addition, to further improve the diversity and precision of features, data enhancement technology is used. Data enhancement generates different training samples by rotating, translating, flipping, and other operations on the original image, which increases the diversity of training data and helps the network learn more robust features. Through these means, the most representative information is extracted from medical images, providing strong support for subsequent efficacy evaluation, drug safety monitoring, and pricing optimization. The overall process is shown in Fig. 2.
Fig. 2 [Images not available. See PDF.]
CNN processing of lesion images.
Efficacy evaluation and safety monitoring
In the research and development of children’s drugs, it is crucial to ensure the efficacy and safety of drugs. This study evaluates the efficacy and safety of drugs through medical imaging data, and uses the image difference method to compare the volume changes of tumors or lesion areas before and after treatment to quantify the drug efficacy. At the same time, the U-Net network is used for image segmentation to precisely extract the boundaries and morphological changes of the lesion area to improve the accuracy of efficacy evaluation. For safety evaluation, imaging technology monitors the changes of other important organs of children, analyzes the degree of organ damage through volume change rate, and quantifies side effects. Finally, regression analysis is used to comprehensively evaluate efficacy and safety to provide support for R&D decisions. To improve the evaluation accuracy, a hierarchical algorithm is used to precisely calibrate the tumor boundary, and the accuracy of the image difference method is further verified by comparing with the expert annotation results25,26. For multi-scale and multi-morphological lesion areas, this study constructs a staged image analysis system. In the image processing of children’s brain tumors, the structural similarity index optimization algorithm is used to enhance the local feature contrast in order to address the problem that small lesions are easily disturbed by noise. The medical image segmentation system based on the U-Net architecture achieves millimeter-level extraction of lesion boundaries through an encoder-decoder bidirectional path. Among them, the encoder implements feature dimensionality reduction; the decoder performs spatial information recovery; the cross-level jump connection effectively integrates shallow texture and deep semantic features. This architecture maintains the integrity of the lesion morphology while achieving quantitative tracking of lesion volume changes during the treatment cycle.
The CNN feature extraction model is validated on an independent test set (1,075 cases), and the performance of the segmentation task is evaluated using the Dice coefficient (0.89 ± 0.03), sensitivity (92.1%), and specificity (94.7%); the U-Net lesion segmentation model achieves a Dice coefficient of 0.91 ± 0.02 on brain tumor data. Training hyperparameters: learning rate 0.001 (Adam optimizer), batch size 32, iteration 200 rounds, and early stopping strategy (tolerance = 10 rounds) are used to prevent overfitting.
The U-Net encoder uses 4 levels of downsampling (3 × 3 convolution kernel); the decoder uses transposed convolution and skip connection; the final output layer generates a binary mask through Sigmoid activation. CNN is based on the ResNet-50 pre-trained model, freezing the weights of the first 20 layers, and then globally average pooling and connecting to the fully connected layer (Dropout = 0.5). Training takes 72 h (NVIDIA V100 GPU).
In terms of network configuration, the input layer receives 256 × 256 pixel DICOM format images; the convolution kernel uses a 3 × 3 size to ensure local feature capture capabilities; the maximum pooling layer (2 × 2) controls the feature map downsampling rate. Data enhancement strategies such as random rotation (± 15°), horizontal flipping, and translation (± 10%) are applied during the training phase to significantly improve the model’s adaptability to body position differences.
The segmentation results directly serve the treatment evaluation system, and by extracting the boundaries and morphological changes of the lesion area, the inhibition rate of targeted drugs on lesions is quantitatively analyzed. For safety evaluation, this study uses imaging technology to monitor the volume changes of important organs (such as the heart and liver) in children and quantify the possible side effects of drugs27,28. By precisely registering images before and after treatment, image segmentation is used to calculate the organ volume change rate and evaluate the effect of drugs on organs. At the same time, multi-time point image data is used to improve the evaluation accuracy. Regression analysis is used to quantify the relationship between efficacy and safety, identify key factors affecting efficacy and side effects, and provide a basis for drug optimization. The effects of different drug ingredients and treatment regimens on efficacy and safety are accurately evaluated through regression analysis, providing data support for drug research and development29,30.
Fig. 3 [Images not available. See PDF.]
Effect of drug ingredients on efficacy and safety.
In Fig. 3, drug ingredients have a significant impact on efficacy and safety. As the drug dose increases, the efficacy is significantly improved, while the safety score decreases. In addition, the effect of ingredients on efficacy and safety is relatively gentle. The effect of treatment time relative to the dose is small. As the treatment time increases, the efficacy is improved to a certain extent, but there is also a certain risk of side effects. These results provide important data support for drug R&D and help optimize drug formulations.
Design of R&D incentive mechanism
In the process of children’s drug R&D, pharmaceutical companies face the challenges of high risk and low return, resulting in insufficient investment. To this end, this study designs an R&D incentive mechanism based on imaging data, quantifies the efficacy and safety of drugs, and establishes an efficacy-safety evaluation model, providing data support for the incentive mechanism. The model evaluates the efficacy and side effects of drugs by analyzing the volume change of the lesion area and the organ volume change rate before and after treatment. Based on the evaluation results, a reasonable profit distribution plan is designed in combination with game theory to encourage companies to increase R&D investment. At the same time, Monte Carlo simulation is used to evaluate R&D risks and provide decision-making basis for the company. The comprehensive evaluation of efficacy and safety is carried out by weighted summation, and the weight coefficient is combined with regression analysis and expert opinions to ensure the model’s scientificity and rationality, thereby providing support for the design of incentive mechanisms31,32.
In drug research and development, pharmaceutical companies face the problem of high costs and low returns, especially insufficient investment in children’s drug research and development. To solve this problem, this study uses a game theory model to design a reasonable R&D incentive mechanism and analyzes the strategic choices of pharmaceutical companies in R&D investment and profit distribution. The drug market value is determined by the efficacy-safety evaluation model, and an R&D incentive plan is formulated. The reward mechanism and profit distribution rules are designed through the Nash equilibrium model to encourage pharmaceutical companies to increase R&D investment in children’s drugs and ensure the effective allocation of resources. “Incentive mechanism” specifically refers to the direct cash subsidies provided by the government in proportion to R&D investment.
Table 1. Incentive mechanism and profit distribution in children’s drug research and development note: the currency units in the table are all in millions of U.S. Dollars (USD), which are consistent with the pricing model units in the following text.
R&D investment (in millions) | Drug market value (in millions) | R&D cost (in millions) | R&D duration (in months) | Profit distribution ratio (Company/Government) | Incentive mechanism (in millions) | Nash equilibrium profit (in millions) |
|---|---|---|---|---|---|---|
5 | 50 | 18 | 12 | 70% / 30% | 3 | 28 |
10 | 90 | 30 | 18 | 65% / 35% | 6 | 45 |
15 | 150 | 45 | 24 | 60% / 40% | 12 | 65 |
20 | 200 | 65 | 27 | 55% / 45% | 18 | 85 |
25 | 280 | 90 | 30 | 50% / 50% | 25 | 110 |
Table 1 analyzes the incentive mechanism and profit distribution of pharmaceutical companies in children’s drug research and development through a game theory model. With the increase in R&D investment, the drug market value and R&D cost are on the rise. An investment of 5 million corresponds to a market value of 50 million and a Nash equilibrium profit of 28 million, while an investment of 25 million corresponds to a market value of 280 million and a profit of 110 million. The government’s share in the profit distribution also increases from 30% to 50% with the increase in investment, indicating that the government has increased its efforts in incentivizing research and development. Meanwhile, the reward mechanism also increases with investment. These data ensure the market accessibility and social benefits of drugs.
In this game theory model, pharmaceutical companies must not only consider their own R&D investment, but also the strategies of market competitors. Therefore, this study also adopts dynamic game theory, taking into account the uncertainty and market changes in the drug R&D process, and designs a dynamically adjusted R&D incentive mechanism. By continuously adjusting the profit distribution plan and reward mechanism, the pharmaceutical company can achieve the best strategic choice at different stages and maximize the R&D investment and the success rate of drug R&D. In the dynamic game model, the pharmaceutical company adjusts its R&D investment according to market changes and the strategies of other competitors. Based on factors such as market demand and R&D cycle, the profit distribution plan in the dynamic game can be expressed as Formula (1):
1
Among them, is the profit of pharmaceutical company i at time point t. is the incentive coefficient of company i at time t (related to the incentive mechanism). is the market value of the drug calculated based on company i’s R&D investment , market demand , and competitor’s strategy . is the R&D cost of company i at time point t. is the influence coefficient of company i on company j’s R&D investment, which is used to reflect the interaction between competitors. The entire dynamic game model is shown in Fig. 4.
Fig. 4 [Images not available. See PDF.]
Dynamic game model.
Risk management is an important link in the process of drug R&D, especially in the R&D of children’s drugs, where the side effects and uncertainty of drug efficacy may bring great risks to companies. To effectively evaluate the risks in the process of drug R&D, this study uses the Monte Carlo simulation method to systematically evaluate the risks that may be encountered in the process of R&D. Monte Carlo simulation models various uncertain factors involved in the process of drug R&D through random sampling and probability analysis, such as drug efficacy, R&D costs, market demand fluctuations, etc.
This study uses Monte Carlo simulation technology to construct a drug R&D risk assessment system. Based on the efficacy-safety evaluation model framework, a multivariate probability distribution model is established to simulate the dynamic evolution of drug efficacy indicators in phase I-III clinical trials, and at the same time integrate the adverse reaction incidence database of similar drugs in the past to quantify the safety risk thresholds at each stage. Through random sampling operations, a joint distribution surface of key indicators such as R&D success rate, cost overrun probability, and market return rate is generated, and the comprehensive risk level of new drug R&D is evaluated based on this33,34.
To improve the credibility of the simulation results, the pharmacokinetic parameters in historical clinical trials are used as prior distributions to perform Bayesian calibration on the parameter space of the Monte Carlo model. By applying data from the actual drug R&D process into the simulation, the accuracy of risk assessment is improved. Combined with the results of imaging data analysis, the Monte Carlo simulation provides pharmaceutical companies with a detailed report on drug R&D risks.
In the Monte Carlo simulation process, after multiple random samplings, multiple simulations are performed to obtain multiple possible results in the drug R&D process. The total number of simulations is set to N, and then the total profit of the k-th simulation is Pk. The average value and risk of the simulation results are calculated by Formulas (2) and (3):
2
3
Among them: is the average value of the simulation results, representing the average profit of drug R&D. is the change in efficacy (such as the rate of change in efficacy), which is a variable provided by the efficacy-safety assessment model. is the R&D cost, which may change according to different random sampling results. is market demand, which is usually manifested as fluctuations in demand for drugs. is the probability of drug side effects, which is usually estimated based on historical data. is the time period of the R&D process, which may affect the marketing and risk of drugs. is the result of the k-th simulation, reflecting the specific risks in the drug R&D process.
Nash equilibrium parameters: corporate risk aversion coefficient λ = 0.8 (referring to PhRMA industry report), government subsidy elasticity coefficient η = 1.2.
Monte Carlo settings: 10,000 iterations. The prior distribution uses Beta (α = 1.2, β = 2.6) to fit the historical R&D success rate, and the sensitivity analysis shows that when the demand volatility σ > 30%, the profit risk surges (VaR > 25%). Verification: compared with the linear regression baseline model, the risk prediction error of this solution is reduced by 19.3%.
Construction of pricing optimization model
Based on medical imaging data, this study constructs a pricing optimization model for children’s drugs, aiming to balance drug efficacy and safety, reasonably set market prices, and improve the market competitiveness of drugs. The model comprehensively considers factors such as production costs, R&D investment, market demand, competitor pricing, and drug effects, and adopts linear programming and dynamic pricing strategies to ensure the economic benefits and social accessibility of drugs. The dynamic pricing model incorporates regional medical policy differences as a key regulator. For low-income countries (such as Africa), the model significantly lowers the price threshold and integrates the WHO essential drug list priority and local government subsidy mechanism (the subsidy ratio can be as high as 40%); while for high-income countries (such as Europe and the United States), the premium space is adjusted by focusing on commercial insurance coverage and the remaining patent protection period. All pricing strategies are calibrated through a regularly updated regional policy database that covers the latest policy trends in 12 representative markets around the world (including China, the United States, the European Union, Africa, etc.).
This study evaluates the efficacy and safety of drugs through medical imaging data to provide basic support for drug pricing models. After quantifying drug efficacy and safety, pricing needs to consider factors such as production costs, R&D investment, and market demand, and provide pricing benchmarks based on clinical value35,36. To optimize pricing, this study adopts a linear programming method to maximize drug profits through an optimization process, considering key factors such as sales volume, production costs, pricing, and market share. Specifically, the objective function can be expressed as:
The profit maximization objective function represents the final revenue of the drug in the market, taking into account factors such as sales volume, unit price, production cost, and market share, and can be expressed as Formula (4):
4
Among them: represents profit; is the drug pricing (price); is the drug production cost; is the drug sales volume, which is a function of price and market share . Market share is usually a function of factors such as pricing, market demand, and competitive pricing. Assuming that market share is related to drug pricing, it can be represented by a linear or nonlinear function. A simple assumption is:
5
Among them: is the drug market share; is a constant (weight) related to market demand; is the total market demand; is the drug pricing; is the sensitivity coefficient of the price to the market share.
The production cost and R&D investment of drugs are usually obtained through accounting and financial data. The model also takes into account the impact of competitor pricing. Assuming that there are multiple competing drugs in the market, the price setting needs to take into account the pricing level of competitors.
Meanwhile, considering that drug pricing is not just a single decision variable, but also related to factors such as patient acceptance, medical insurance reimbursement policies, and subsidies at the policy level, appropriate constraints are added to the model to ensure that pricing is reasonable and in line with social benefits. For example, drug pricing should not exceed the upper limit of medical insurance reimbursement and should maintain a certain market share.
Based on the preliminary pricing model, this study further applies a dynamic pricing strategy. Drug pricing is not only affected by the initial research and development stage, but also by market response, consumer demand fluctuations, and competitor price adjustments. Therefore, the dynamic pricing strategy is to address the risks brought by market uncertainty and maintain the flexibility and market adaptability of drug pricing.
The core idea of the dynamic pricing strategy is to adjust prices in a timely manner according to market response and drug performance. This study uses a historical sales data prediction model and a market feedback mechanism, combined with imaging data evaluation results, to dynamically adjust drug prices. The specific steps include: (1) market feedback is collected, such as sales data, demand, and competitor pricing, and market performance is monitored; (2) price adjustment thresholds are set according to demand changes to reduce prices when demand is low and increase prices when demand is high; (3) a reasonable adjustment cycle is set, such as evaluating and adjusting prices every quarter. Through this strategy, it is ensured that the drug is competitive in the early stage and obtains stable profits in the long run. At the same time, the linear programming model and market feedback are combined to optimize the pricing range, ensure the economic benefits and social accessibility of the drug, and adjust the pricing in time to respond to market changes.
Interdisciplinary Cooperation and policy recommendations
This study emphasizes the collaboration of interdisciplinary teams and forms teams of medical imaging experts, drug R&D experts, and market analysts, combining the advantages of each discipline to promote the application of medical imaging technology in children’s drug research and development and pricing. The team’s goal is to combine medical imaging with the drug research and development process to provide precise efficacy and safety data, thereby optimizing drug pricing strategies.
In this process, medical imaging experts are responsible for providing technical support for the collection and analysis of imaging data to ensure the accuracy and effectiveness of the data. Drug R&D experts use imaging data to evaluate the performance of different drugs in children, analyze the efficacy and safety of drugs, and adjust the direction of drug research and development based on the feedback of imaging data. Market analysts combine market demand, drug pricing, competitive environment, and other factors to provide decision-making basis for drug pricing. At the same time, experts from various disciplines share each other’s research progress through regular meetings and workshops, solve technical and theoretical problems in practical applications, and promote data-driven drug R&D processes.
The collaboration of the interdisciplinary team not only provides multi-angle insights in the process of data analysis and drug R&D, but also promotes the optimization of drug pricing strategies. By integrating medical imaging analysis, drug benefit evaluation, and market demand prediction, a pricing model that is more in line with market demand is designed, providing strong theoretical support for drug pricing.
Based on the research results of the interdisciplinary team, this study puts forward data-driven policy recommendations to promote the government to formulate policies that are conducive to the research and development of children’s drugs. By analyzing the application of medical imaging technology in drug R&D, its potential to improve R&D efficiency and safety is evaluated, and specific policy recommendations are provided to the government, focusing on R&D subsidies and tax incentives. It is recommended that the government sets up a special fund to support imaging data-driven drug R&D, especially children’s drugs, to reduce the R&D burden of companies. At the same time, it is recommended to refer to imaging data in drug pricing policies to ensure scientific and reasonable pricing and to ensure the competitiveness and accessibility of drugs in the market. Interdisciplinary collaboration and policy support are the key to promoting the optimization of children’s drug R&D and market pricing.
Evaluation indicators and calculation methods
In the process of drug R&D and pricing optimization, reasonable evaluation indicators and calculation methods are of great significance for the comprehensive evaluation of drug efficacy, safety, and the rationality of market pricing. This study designs corresponding evaluation indicators from three main aspects: drug efficacy, safety, and market pricing, and uses a variety of calculation methods to ensure the scientificity and accuracy of the evaluation results.
Drug efficacy indicators
The core of drug efficacy evaluation is to quantify the effect of drugs on target areas (tumors, bone density, etc.) through medical imaging data. In this study, the imaging data before and after treatment is first used to analyze the imaging differences and calculate the tumor volume and bone density change rate. To achieve this goal, the image difference method is used to register the imaging data at different time points, and the percentage change of tumor volume and bone density before and after drug treatment is calculated. This method determines the drug efficacy by comparing the imaging data at different time points.
Fig. 5 [Images not available. See PDF.]
Volume change trend of five tumors during treatment.
Figure 5 shows the volume change trend of five tumors during treatment. Tumors A and B decrease rapidly at the beginning of treatment and finally stabilize at 29.1 cm³ and 28 cm³, indicating that the drug is effective in the early stage and gradually weakens after recovery. The volume of tumor C decreases from 50 cm³ to 40.0 cm³, and the volume of tumors D and E also decrease significantly. After multiple experimental evaluations, the average tumor volume decreases by 32.7% (95% CI: 28.4%-36.9%). These data reflect the quantitative effect change through medical imaging, providing an important basis for the R&D of children’s drugs.
Fig. 6 [Images not available. See PDF.]
Recovery of bone density of patients after treatment.
Figure 6 shows the recovery of bone density of 5 patients after treatment. The bone density of all patients gradually recovers and stabilizes. For example, patient 1 recovers from 0.82 g/cm³ to 0.94 g/cm³, showing continuous improvement and final stabilization; patient 3 recovers well and stabilizes at 0.97 g/cm³. The recovery of other patients, such as patients 2 and 5, eventually stabilizes at 0.89 g/cm³ and 0.86 g/cm³, respectively. This shows that the treatment has a significant effect on bone density, and the individual recovery is good.
Safety indicators
The safety assessment of drugs is an important link that cannot be ignored in the research and development process, especially in the R&D of children’s drugs, where its impact is particularly significant. This study uses imaging data to monitor the effects of drugs on children’s organs (such as heart, liver, etc.), and specifically uses volume change rate and morphological change indicators to quantitatively analyze the side effects of drugs. The volume change rate reflects the effect of drug treatment on organ volume. If the drug causes an increase or decrease in organ volume, it may indicate potential side effects. Imaging follow-up is strictly carried out in accordance with the clinical trial protocol, at baseline (T0) and once a month after treatment (T1-T9), with an interval of 30 ± 2 days (CT/MRI layer thickness 1 mm). Heart volume monitoring uses a dynamic MRI short-axis sequence (time resolution 40ms) to ensure that instantaneous changes are captured.
This study extracts organ contours through image segmentation, calculates morphological changes, and quantitatively analyzes the effects of drugs on organs, providing data support for safety assessment. At the same time, combined with multivariate regression analysis, the relationship between drug ingredients and children’s organ damage is explored, and the key factors affecting drug efficacy and side effects are identified. Tables 2 and 3 show the effects of drugs A and B on the volume of children’s heart and liver, reflecting the changes in organ volume during the treatment of the two drugs.
Table 2. Effects of drugs A and B on heart volume.
Time Point | Drug a heart volume (cm3) | Drug B Heart Volume (cm3) |
|---|---|---|
0 | 85 | 80 |
1 | 85.1 | 80.1 |
2 | 84.9 | 79.8 |
3 | 85.2 | 80 |
4 | 85 | 79.9 |
5 | 85.1 | 80.2 |
6 | 84.8 | 80 |
7 | 85 | 79.8 |
8 | 85.2 | 80.1 |
9 | 85.1 | 80 |
The data in Table 2 shows that patients taking drug A have a smaller effect on heart volume, with a heart volume of 85 cm³ at time 0, and a smaller change range thereafter, fluctuating between 84.8 cm³ and 85.2 cm³, showing the stability of heart volume. Drug B also has a smaller effect on heart volume, with a starting volume of 80 cm³ and a fluctuation range of 79.8 cm³ and 80.2 cm³, with smaller fluctuations and less impact on the heart.
Table 3. Effects of drugs A and B on liver volume.
Time point | Drug A liver volume (cm3) | Drug B liver volume (cm3) |
|---|---|---|
0 | 110 | 105 |
1 | 112 | 105.7 |
2 | 111 | 105.8 |
3 | 110.5 | 104.9 |
4 | 110.5 | 105.6 |
5 | 110.6 | 104.8 |
6 | 109 | 105 |
7 | 109.5 | 105.2 |
8 | 109.9 | 105.3 |
9 | 110.1 | 105.4 |
Table 3 shows that the liver volume of patients taking drug A varies around 110 cm2, with a maximum fluctuation of 2 cm2, which is within the normal range. For patients taking drug B, the liver volume varies around 105 cm2. These data show that both drugs A and B have very small effects on the volume of the heart and liver during treatment, indicating that these two drugs may have little side effects on these organs when used clinically, and long-term use of the drugs does not lead to significant changes in organ volume.
The threshold of organ volume change of ± 2 cm2 is based on two clinical criteria: (1) the Pediatric Imaging Consensus Statement stipulates that liver/heart volume fluctuations of > 5% require intervention. Combined with the average organ volume of the cases in this study, ± 2 cm³2is the normal value ; (2) the WHO drug safety guidelines list a volume change of < 3% as an “acceptable risk.”
Rationality of market pricing
The rationality of drug pricing is directly related to its market acceptance and competitiveness. This study evaluates the rationality of drug pricing by constructing a regression model. The regression model takes into account multiple key variables, including drug production costs, R&D investment, market demand, and competitor pricing. By weighting these variables, the reasonable pricing range of drugs is calculated. In particular, drug production costs and R&D investment are the main factors determining pricing, while market demand and competitive situation further affect the market positioning and pricing strategy of drugs.
This study evaluates the impact of different pricing strategies on sales and market share through price elasticity analysis. Price elasticity reflects the sensitivity of price changes to sales and helps determine the optimal pricing range. This analysis not only provides pharmaceutical companies with the best pricing point, but also allows policymakers to consider market demand elasticity when formulating drug pricing policies and improve drug competitiveness.
Fig. 7 [Images not available. See PDF.]
Analysis of drug pricing and market sales.
Figure 7 shows the nonlinear relationship between drug pricing and market sales and price elasticity analysis. In the left figure, price and sales show an obvious power law relationship. As drug prices increase, sales drop sharply. The optimal pricing point is set at $120, and overall sales reach the highest. The green dot indicates this optimal price. The right figure shows the price elasticity analysis, indicating that as prices increase, price elasticity gradually decreases, and the sensitivity of price changes to sales decreases. At the optimal pricing point ($120), the price elasticity is -1.338, indicating that a small adjustment in price has a significant impact on sales. If prices are further increased, the reduction in sales accelerates. These data show that by reasonably analyzing price elasticity, pharmaceutical companies can achieve maximum market share and profits while ensuring competitiveness.
Through the above evaluation indicators and calculation methods, this paper comprehensively analyzes the efficacy, safety, and rationality of market pricing of drugs, and provides a scientific basis for drug R&D and market strategies. The implementation of these evaluation methods not only helps to optimize the drug R&D process, but also provides precise quantitative support for the market pricing of children’s drugs, promoting their competitiveness and accessibility in the market.
Conclusions
Based on medical imaging technology, this paper proposes a data-driven incentive mechanism for pediatric drug development and a market pricing optimization method. This exploratory study proposes a framework that combines image quantification and economic models in a simulation verification environment. All efficacy and pricing conclusions have not been verified by real-world clinical endpoints. This study combines CNN and U-Net image segmentation technology to evaluate the efficacy and safety of pediatric drugs, uses imaging data to monitor efficacy and side effects, and designs a development incentive mechanism through game theory, combining linear programming and dynamic pricing to optimize market pricing. A framework that integrates image quantification and economic models in a simulation environment provides a preliminary reference for optimizing the pediatric drug development process and improving pricing rationality. However, this paper still has some limitations, especially in terms of diversity and sample size in data collection and processing : (1) Data source: 87% of the imaging data comes from retrospective clinical trials, and the rest are prospective observation cohorts (insufficient sample size may introduce bias); (2) Economic model verification: Monte Carlo simulation is based on historical market data and has not yet been stress-tested for sudden policy changes. In addition, although the proposed pricing optimization model is highly operational, in actual application, changes in the market and policy environment may also affect the model’s effectiveness. Future research can consider further enriching data sources, optimizing models, and combining artificial intelligence and big data technologies to improve the intelligence and adaptability of drug development and market pricing.
Author contributions
All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Ethical approval
This research was approved by the Ethics Committee of China Pharmaceutical University. I confirm that all methods were performed in accordance with the relevant guidelines.
Informed consent
I confirming that informed consent was obtained from all subjects and/or their legal guardians.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
1. Greenberg, RG et al. Pediatric clinical research networks: role in accelerating development of therapeutics in children. Therapeutic Innov. Regul. Sci.; 2022; 56,
2. Wakutsu, N et al. Assessing definitions and incentives adopted for innovation for pharmaceutical products in five high-income countries: a systematic literature review. Pharm. Med.; 2023; 37,
3. Annett, S. Pharmaceutical drug development: high drug prices and the hidden role of public funding. Biol. Futura; 2021; 72,
4. Beckman, RA et al. Innovations in clinical development in rare diseases of children and adults: small populations and/or small patients. Pediatr. Drugs; 2022; 24,
5. Lo, AW; Thakor, RT. Financing biomedical innovation. Annual Rev. Financial Econ.; 2022; 14,
6. Qiu, T et al. Challenges in the market access of regenerative medicines, and implications for manufacturers and decision-makers: a systematic review. Regen. Med.; 2022; 17,
7. Chan, AYL et al. Access and unmet needs of orphan drugs in 194 countries and 6 areas: a global policy review with content analysis. Value Health; 2020; 23,
8. Shah, KK; Kogut, S; Slitt, A. Challenges in evaluating safety and efficacy in drug development for rare diseases: a review for pharmacists. J. Pharm. Pract.; 2021; 34,
9. Idoko, IP et al. The dual impact of AI and renewable energy in enhancing medicine for better diagnostics, drug discovery, and public health. Magna Scientia Adv. Biology Pharm.; 2024; 12,
10. Roessler, HI et al. Drug repurposing for rare diseases. Trends Pharmacol. Sci.; 2021; 42,
11. Beall, RF et al. Reimagining pharmaceutical market exclusivities: should the duration of guaranteed monopoly periods be value based?. Value Health; 2021; 24,
12. Irissarry, C; Burger-Helmchen, T. Using artificial intelligence to advance the research and development of orphan drugs. Businesses; 2024; 4,
13. Tu, Q et al. Individualized precision dosing approaches to optimize antimicrobial therapy in pediatric populations. Expert Rev. Clin. Pharmacol.; 2021; 14,
14. Gonçalves, E. Value-based pricing for advanced therapy medicinal products: emerging affordability solutions. Eur. J. Health Econ.; 2022; 23,
15. Crossnohere, NL et al. The evolution of patient-focused drug development and Duchenne muscular dystrophy. Expert Rev. PharmacoEcon. Outcomes Res.; 2020; 20,
16. Liu, Y et al. Evolution of drug regulations and regulatory innovation for anticancer drugs in China. Acta Pharm. Sinica B; 2022; 12,
17. Xu, Y et al. Exploring patient medication adherence and data mining methods in clinical big data: A contemporary review. J. Evidence-Based Med.; 2023; 16,
18. Miera-Maluenda, M et al. Advances in the approaches used to repurpose drugs for neuroblastoma. Expert Opin. Drug Discov.; 2024; 19,
19. Wan, M et al. Access to child-appropriate medicines: an exploratory survey of the use of paediatric use marketing authorisation products in the UK. Eur. J. Pediatrics; 2025; 184,
20. Darwich, AS et al. Model-informed precision dosing: background, requirements, validation, implementation, and forward trajectory of individualizing drug therapy. Annu. Rev. Pharmacol. Toxicol.; 2021; 61,
21. Mishra, AS; Vasanthan, M; Malliappan, SP. Drug repurposing: A leading strategy for new threats and targets. ACS Pharmacol. Translational Sci.; 2024; 7,
22. Godman, B; McCabe, H; Leong, D. Fixed dose drug combinations–are they pharmacoeconomically sound? Findings and implications especially for lower-and middle-income countries. Expert Rev. PharmacoEcon. Outcomes Res.; 2020; 20,
23. Page, S et al. Patient centricity driving formulation innovation: improvements in patient care facilitated by novel therapeutics and drug delivery technologies. Annu. Rev. Pharmacol. Toxicol.; 2022; 62,
24. Gaudilliere, JP. Pharmaceutical innovation and its crisis: drug markets, screening, and the dialectics of value. BioSocieties; 2021; 16,
25. Landfeldt, E. Gene therapy for neuromuscular diseases: health economic challenges and future perspectives. J. Neuromuscul. Dis.; 2022; 9,
26. Petyko, ZI et al. Barriers and facilitators of exploiting the potential of value-added medicines. Expert Rev. PharmacoEcon. Outcomes Res.; 2020; 20,
27. Dutta, B et al. Identifying key benefits in European off-patent biologics and biosimilar markets: it is not only about price!. BioDrugs; 2020; 34,
28. Qiu, T et al. Regenerative medicine regulatory policies: a systematic review and international comparison. Health Policy; 2020; 124,
29. Caferra, P et al. Evaluation of orphan maintained biological medicinal products in the European union between 2018 to 2023: a regulatory perspective. Expert Opin. Biol. Ther.; 2024; 24,
30. Saha, D; Vasuprada, TM. Investigating commercial incentives for innovation: an application in traditional medicine. Stud. Microeconomics; 2021; 9,
31. Ye, J; Bi, Y; Ting, N. How to select the initial dose for a pediatric study?. J. Biopharm. Stat.; 2023; 33,
32. Cruz, ED et al. US food and drug administration facilitated pediatric approval programs: application to pediatric neurological disorders. J. Child. Neurol.; 2022; 37,
33. Piddock, LJV et al. Advancing global antibiotic research, development and access. Nat. Med.; 2024; 30,
34. Bubela, T et al. Open drug discovery of anti-virals critical for canada’s pandemic strategy. Facets; 2020; 5,
35. Fontrier, AM; Kanavos, P. Do reimbursement recommendations by the Canadian agency for drugs and technology in health translate into coverage decisions for orphan drugs in the Canadian Province of ontario?. Value Health; 2023; 26,
36. Rodwin, MA. How the united Kingdom controls pharmaceutical prices and spending: learning from its experience. Int. J. Health Serv.; 2021; 51,
© The Author(s) 2025. This work is published under http://creativecommons.org/licenses/by-nc-nd/4.0/ (the "License"). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.