A technique for using neural network analysis to perform survival analysis of censored data

Advancing healthcare: the role and impact of AI and foundation models

BACKGROUND

The integration of artificial intelligence (AI) into the healthcare domain is a monumental shift with profound implications for diagnostics, medical interventions, and the overall structure of healthcare systems.

PURPOSE

This study explores the transformative journey of foundation AI models in healthcare, shedding light on the challenges, ethical considerations, and vast potential they hold for improving patient outcome and system efficiency. Notably, in this investigation we observe a relatively slow adoption of AI within the public sector of healthcare. The evolution of AI in healthcare is un-paralleled, especially its prowess in revolutionizing diagnostic processes.

RESULTS

This research showcases how these foundational models can unravel hidden patterns within complex medical datasets. The impact of AI reverberates through medical interventions, encompassing pathology, imaging, genomics, and personalized healthcare, positioning AI as a cornerstone in the quest for precision medicine. The paper delves into the applications of generative AI models in critical facets of healthcare, including decision support, medical imaging, and the prediction of protein structures. The study meticulously evaluates various AI models, such as transfer learning, RNN, autoencoders, and their roles in the healthcare landscape. A pioneering concept introduced in this exploration is that of General Medical AI (GMAI), advocating for the development of reusable and flexible AI models.

CONCLUSION

The review article discusses how AI can revolutionize healthcare by stressing the significance of transparency, fairness and accountability, in AI applications regarding patient data privacy and biases. By tackling these issues and suggesting a governance structure the article adds to the conversation about AI integration in healthcare environments.

Artificial intelligence, machine learning, and deep learning for clinical outcome prediction

AI is a broad concept, grouping initiatives that use a computer to perform tasks that would usually require a human to complete. AI methods are well suited to predict clinical outcomes. In practice, AI methods can be thought of as functions that learn the outcomes accompanying standardized input data to produce accurate outcome predictions when trialed with new data. Current methods for cleaning, creating, accessing, extracting, augmenting, and representing data for training AI clinical prediction models are well defined. The use of AI to predict clinical outcomes is a dynamic and rapidly evolving arena, with new methods and applications emerging. Extraction or accession of electronic health care records and combining these with patient genetic data is an area of present attention, with tremendous potential for future growth. Machine learning approaches, including decision tree methods of Random Forest and XGBoost, and deep learning techniques including deep multi-layer and recurrent neural networks, afford unique capabilities to accurately create predictions from high dimensional, multimodal data. Furthermore, AI methods are increasing our ability to accurately predict clinical outcomes that previously were difficult to model, including time-dependent and multi-class outcomes. Barriers to robust AI-based clinical outcome model deployment include changing AI product development interfaces, the specificity of regulation requirements, and limitations in ensuring model interpretability, generalizability, and adaptability over time.

A novel diagnostic and prognostic approach for unresponsive patients with anthroponotic cutaneous leishmaniasis using artificial neural networks

Cutaneous leishmaniasis (CL) imposes a major health burden throughout the tropical and subtropical regions of the globe. Unresponsive cases are common phenomena occurred upon exposure to the standard drugs. Therefore, rapid detection, prognosis and classification of the disease are crucial for selecting the proper treatment modality. Using machine learning (ML) techniques, this study aimed to detect unresponsive cases of ACL, caused by Leishmania tropica, which will consequently be used for a more effective treatment modality. This study was conducted as a case-control setting. Patients were selected in a major ACL focus from both unresponsive and responsive cases. Nine unique and relevant features of patients with ACL were selected. To categorize the patients, different classifier models such as k-nearest neighbors (KNN), support vector machines (SVM), multilayer perceptron (MLP), learning vector quantization (LVQ) and multipass LVQ were applied and compared for this supervised learning task. Comparison of the receiver operating characteristic graphs (ROC) and confusion plots for the above models represented that MLP was a fairly accurate prediction model to solve this problem. The overall accuracy in terms of sensitivity, specificity and area under ROC curve (AUC) of MLP classifier were 87.8%, 90.3%, 86% and 0.88%, respectively. Moreover, the duration of the skin lesion was the most influential feature in MLP classifier, while gender was the least. The present investigation demonstrated that MLP model could be utilized for rapid detection, accurate prognosis and effective treatment of unresponsive patients with ACL. The results showed that the major feature affecting the responsiveness to treatments is the duration of the lesion. This novel approach is unique and can be beneficial in developing diagnostic, prophylactic and therapeutic measures against the disease. This attempt could be a preliminary step towards the expansion of ML application in future directions.

Peri-Tumoral Inflammatory Cell Infiltration in OSCC: A Reliable Marker of Local Recurrence and Prognosis? An Investigation Using Artificial Neural Networks

The presence of inflammatory reaction in peri-tumoural connective tissue is generally considered as a defense mechanism against cancer, but inflammation tissue in malignant transformation and early steps of oncogenesis has been recently proven to play a supporting and aggravating role in some carcinomas. Aims of this retrospective study were to evaluate in OSCCs the independent association of peri-tumoral inflammatory infiltrate (PTI) with local recurrence (LR) or survival outcome, and to verify whether PTI can be considered a marker of prognosis. Data from 211 cases of OSCC, only surgically treated between 1990 and 2000, were collected and retrospectively analyzed for PTI and the event LR (5 yrs follow-up at least) by means of univariate-multivariate and neural networks analyses. Patients (mean age 65.3 ± 12.4 yrs, M/F = 2.98) showed presence of PTI in 68.2% (144/211): (+) in 27.0%, (++) in 25.6%, (+++) 15.6%; PTI was found reduced in 24.7% of cases and absent in 7.1%. In overall PTI+ve group (n=144), 66 were TNM Stage I, 33 Stage II, 45 Stage III, none Stage IV. LR (mean 6 ± 4 months) was present in 87/211 (41.2%) patients, of which 43/144 (29.8%) in OSCCs with PTI [23 (+),. 13 (++) and 7 (+++)] vs. 44/67 (65.7%) in OSCC with PTI -/+ or PTI–ve ones. By univariate analysis, PTI+ve cases showed a significant lower risk to have LR (p<0.0001; OR= 0.2297; CI= 0.1277:0.4134) vs PTI -/+ or –ve ones, especially among cases with higher PTI value (+++) (OR= 0.1718; CI= 0.0749:03939). Multivariate analyses (Logit model and neural networks) confirmed the same datum: presence of PTI was an independent predictive variable accounting for a better tumoural outcome without LR (Logit and neural networks values: OR' 0.226; CI= 0.113:0.454; ROC Area = 0.66, respectively). In terms of prognostic significance, elevated PTI was found to have an independent association with the poorest overall survival rate (P = 0.056). Our findings strongly suggest the importance to investigate routinely PTI in OSCCs, as useful marker of tumoral behavior and prognosis, and warrant further studies on its specific cellular nature.

Assessment of gastric cancer survival: using an artificial hierarchical neural network

This study is designed to assess the application of neural networks in comparison to the Kaplan-Meier and Cox proportional hazards model in the survival analysis. Three hundred thirty gastric cancer patients admitted to and surgically treated were assessed and their post-surgical survival was determined. The observed baseline survival was determined with the three methods of Kaplan-Meier product limit estimator, Cox and the neural network and results were compared. Then the binary independent variables were entered into the model. Data were randomly divided into two groups of 165 each to test the models and assess the reproducibility. The Chi-square test and the multiple logistic model were used to ensure the groups were similar and the data was divided randomly. To compare subgroups, we used the log-rank test. In the next step, the probability of survival in different periods was computed based on the training group data using the Cox proportional hazards and a neural network and estimating Cox coefficient values and neural network weights (with 3 nodes in hidden layer). Results were used for predictions in the test group data and these predictions were compared using the Kaplan-Meier product limit estimator as the gold standard. Friedman and Kruskal-Wallis tests were used for comparisons as well. All statistical analyses were performed using SPSS version 11.5, Matlab version 7.2, Statistica version 6.0 and S_PLUS 2000. The significance level was considered 5% (alpha = 0.05). The three methods used showed no significance difference in base survival probabilities. Overall, there was no significant difference among the survival probabilities or the trend of changes in survival probabilities calculated with the three methods, but the 4 year (48th month) and 4.5 year (54th month) survival rates were significantly different with Cox compared to standard and estimated probabilities in the neural network (p < 0.05). Kaplan-Meier and Cox showed almost similar results for the baseline survival probabilities, but results with the neural network were different: higher probabilities up to the 4th year, then comparable with the other two methods. Estimates from Cox proportional hazards and the neural network with three nodes in hidden layer were compared with the estimate from the Kaplan-Meier estimator as the gold standard. Neither comparison showed statistically significant differences. The standard error ratio of the two estimate groups by Cox and the neural network to Kaplan-Meier were no significant differences, it indicated that the neural network was more accurate. Although we do not suggest neural network methods to estimate the baseline survival probability, it seems these models is more accurately estimated as compared with the Cox proportional hazards, especially with today's advanced computer sciences that allow complex calculations. These methods are preferable because they lack the limitations of conventional models and obviate the need for unnecessary assumptions including those related to the proportionality of hazards and linearity.

Improvement of breast cancer relapse prediction in high risk intervals using artificial neural networks

The objective of this study is to compare the predictive accuracy of a neural network (NN) model versus the standard Cox proportional hazard model. Data about the 3811 patients included in this study were collected within the 'El Alamo' Project, the largest dataset on breast cancer (BC) in Spain. The best prognostic model generated by the NN contains as covariates age, tumour size, lymph node status, tumour grade and type of treatment. These same variables were considered as having prognostic significance within the Cox model analysis. Nevertheless, the predictions made by the NN were statistically significant more accurate than those from the Cox model (p < 0.0001). Seven different time intervals were also analyzed to find that the NN predictions were much more accurate than those from the Cox model in particular in the early intervals between 1-10 and 11-20 months, and in the later one considered from 61 months to maximum follow-up time (MFT). Interestingly, these intervals contain regions of high relapse risk that have been observed in different studies and that are also present in the analyzed dataset.

Artificial neural networks for diagnosis and survival prediction in colon cancer

ANNs are nonlinear regression computational devices that have been used for over 45 years in classification and survival prediction in several biomedical systems, including colon cancer. Described in this article is the theory behind the three-layer free forward artificial neural networks with backpropagation error, which is widely used in biomedical fields, and a methodological approach to its application for cancer research, as exemplified by colon cancer. Review of the literature shows that applications of these networks have improved the accuracy of colon cancer classification and survival prediction when compared to other statistical or clinicopathological methods. Accuracy, however, must be exercised when designing, using and publishing biomedical results employing machine-learning devices such as ANNs in worldwide literature in order to enhance confidence in the quality and reliability of reported data.

Artificial neural networks as prediction tools in the critically ill

The past 25 years have witnessed the development of improved tools with which to predict short-term and long-term outcomes after critical illness. The general paradigm for constructing the best known tools has been the logistic regression model. Recently, a variety of alternative tools, such as artificial neural networks, have been proposed, with claims of improved performance over more traditional models in particular settings. However, these newer methods have yet to demonstrate their practicality and usefulness within the context of predicting outcomes in the critically ill.

Use of artificial neural networks to predict biological outcomes for patients receiving radical radiotherapy of the prostate

BACKGROUND AND PURPOSE

This paper discusses the application of artificial neural networks (ANN) in predicting biological outcomes following prostate radiotherapy. A number of model-based methods have been developed to correlate the dose distributions calculated for a patient receiving radiotherapy and the radiobiological effect this will produce. Most widely used are the normal tissue complication probability and tumour control probability models. An alternative method for predicting specific examples of tumour control and normal tissue complications is to use an ANN. One of the advantages of this method is that there is no need for a priori information regarding the relationship between the data being correlated.

PATIENTS AND METHODS

A set of retrospective clinical data from patients who received radical prostate radiotherapy was used to train ANNs to predict specific biological outcomes by learning the relationship between the treatment plan prescription, dose distribution and the corresponding biological effect. The dose and volume were included as a differential dose-volume histogram in order to provide a holistic description of the available data.

RESULTS

It was shown that the ANNs were able to predict biochemical control and specific bladder and rectum complications with sensitivity and specificity of above 55% when the outcomes were dichotomised. It was also possible to analyse information from the ANNs to investigate the effect of individual treatment parameters on the outcome.

CONCLUSION

ANNs have been shown to learn something of the complex relationship between treatment parameters and outcome which, if developed further, may prove to be a useful tool in predicting biological outcomes.

Evaluation of inherent performance of intelligent medical decision support systems: utilising neural networks as an example

Researchers who design intelligent systems for medical decision support, are aware of the need for response to real clinical issues, in particular the need to address the specific ethical problems that the medical domain has in using black boxes. This means such intelligent systems have to be thoroughly evaluated, for acceptability. Attempts at compliance, however, are hampered by lack of guidelines. This paper addresses the issue of inherent performance evaluation, which researchers have addressed in part, but a Medline search, using neural networks as an example of intelligent systems, indicated that only about 12.5% evaluated inherent performance adequately. This paper aims to address this issue by concentrating on the possible evaluation methodology, giving a framework and specific suggestions for each type of classification problem. This should allow the developers of intelligent systems to produce evidence of a sufficiency of output performance evaluation.

Genetic adaptive neural network to predict biochemical failure after radical prostatectomy: a multi-institutional study

BACKGROUND AND PURPOSE

Despite many new procedures, radical prostatectomy remains one of the commonest methods of treating clinically localized prostate cancer. Both from the physician's and the patient's point of view, it is important to have objective estimation of the likelihood of recurrence, which forms the foundation for treatment selection for an individual patient. Currently, it is difficult to predict the probability of biochemical recurrence (rising serum prostate specific antigen [PSA] concentration) in an individual patient, and approximately 30% of the patients do experience recurrence. Tools predicting the recurrence will be of immense practical utility in the treatment selection and planning follow up. We have utilized preoperative parameters through a computer based genetic adaptive neural network model to predict recurrence in such patients, which can help primary care physicians and urologists in making management recommendations.

PATIENTS AND METHODS

Fourteen hundred patients who underwent radical prostatectomy at participating institutions form the subjects of this study. Demographic data such as age, race, preoperative PSA, systemic biopsy based staging and Gleason scores were used to construct a neural network model. This model simulated the functioning of a trained human mind and learned from the database. Once trained, it was used to predict the outcomes in new patients.

RESULTS

The patients in this comprehensive database were representative of the average prostate cancer patients as seen in USA. Their mean age was 68.4 years, the mean PSA concentration before surgery was 11.6 ng/mL, and 67% patients had a Gleason sum of 5 to 7. The mean length of follow-up was 41.5 months. Eighty percent of the cancers were clinical stage T2 and 5% T3. In our series, 64% of patients had pathologically organ-confined cancer, 33% positive margins, and 14% had seminal vesicle invasion. Lymph node positive patients were not included in this series. Progression as judged by serum PSA was noted in 30.6%. With entry of a few routinely used parameters, the model could correctly predict recurrence in 76% of the patients in the validation set. The area under the curve was 0.831. The sensitivity was 85%, the specificity 74%, the positive predictive value 77%, and the negative predictive value of 83%.

CONCLUSION

It was possible to predict PSA recurrence with a high accuracy (76%). Physicians desiring objective treatment counseling can use this model, and significant cost savings are anticipated because of appropriate treatment selection and patient-specific follow-up protocols. This technology can be extended to other treatments such as watchful waiting, external-beam radiation, and brachytherapy.

A review of evidence of health benefit from artificial neural networks in medical intervention

The purpose of this review is to assess the evidence of healthcare benefits involving the application of artificial neural networks to the clinical functions of diagnosis, prognosis and survival analysis, in the medical domains of oncology, critical care and cardiovascular medicine. The primary source of publications is PUBMED listings under Randomised Controlled Trials and Clinical Trials. The rĵle of neural networks is introduced within the context of advances in medical decision support arising from parallel developments in statistics and artificial intelligence. This is followed by a survey of published Randomised Controlled Trials and Clinical Trials, leading to recommendations for good practice in the design and evaluation of neural networks for use in medical intervention.

A neural network approach to the outcome definition on first treatment with sertraline in a psychiatric population

Therapy decision is one of the most important tasks clinicians have to perform in their clinical practice. The decision process requires taking into account many different factors. The Authors have proposed a neural computing approach for supporting clinical decision analysis. The mathematical model of artificial neural network (ANN) has been applied on a pool of clinical information gathered through case description freely filled by senior psychiatrists into 416 clinical charts. Sertraline, as drug for treatment, has been chosen since its clinical uses range from treatment of depression to that of many other psychiatric clinical conditions so that it has been thought to be a good candidate to this type of study. The ANN performance in forecasting successful and unsuccessful treatment cases showed an overall accuracy of classification of 97.35%. This result suggests a possible future application of this method to obtain a reliable prediction of a given psychiatric patient outcome during a specific psychopharmacological therapy, optimising the decisional making process.

A neural network approach to the diagnosis of morbidity outcomes in trauma care

This paper introduces the application of artificial neural networks to trauma complications assessment. The potential financial benefits of improving on trauma center diagnostic specificity in complications assessment are illustrated and the operational feasibility of the use of diagnostic neural models across institutions is discussed. A prototype neural network model is described, which, after training, succeeds in diagnosing the complication of sepsis in victims of traumatic blunt injury. Its diagnostic performance with 100% sensitivity and 96.5% specificity is accomplished with test data from a regional trauma center. The model is further shown to have correctly detected, during training, incorrectly coded data. The potential this suggests, for parsimonious database scrubbing through the use of neural network models, is discussed.

Experiments to determine whether recursive partitioning (CART) or an artificial neural network overcomes theoretical limitations of Cox proportional hazards regression

New computationally intensive tools for medical survival analyses include recursive patitioning (also called CART) and artificial neural networks. A challenge that remains is to better understand the behavior of these techniques in effort to know when they will be effective tools. Theoretically they may overcome limitations of the traditional multivariable survival technique, the Cox proportional hazards regression model. Experiments were designed to test whether the new tools would, in practice, overcome these limitations. Two datasets in which theory suggests CART and the neural network should outperform the Cox model were selected. The first was a published leukemia dataset manipulated to have a strong interaction that CART should detect. The second was a published cirrhosis dataset with pronounced nonlinear effects that a neural network should fit. Repeated sampling of 50 training and testing subsets was applied to each technique. The concordance index C was calculated as a measure of predictive accuracy by each technique on the testing dataset. In the interaction dataset, CART outperformed Cox (P < 0.05) with a C improvement of 0.1 (95% CI, 0.08 to 0.12). In the nonlinear dataset, the neural network outperformed the Cox model (P < 0.05), but by a very slight amount (0.015). As predicted by theory, CART and the neural network were able to overcome limitations of the Cox model. Experiments like these are important to increase our understanding of when one of these new techniques will outperform the standard Cox model. Further research is necessary to predict which technique will do best a priori and to assess the magnitude of superiority.

Artificial neural networks applied to outcome prediction for colorectal cancer patients in separate institutions

BACKGROUND

Artificial neural networks are computer programs that can be used to discover complex relations within data sets. They permit the recognition of patterns in complex biological data sets that cannot be detected with conventional linear statistical analysis. One such complex problem is the prediction of outcome for individual patients treated for colorectal cancer. Predictions of outcome in such patients have traditionally been based on population statistics. However, these predictions have little meaning for the individual patient. We report the training of neural networks to predict outcome for individual patients from one institution and their predictive performance on data from a different institution in another region.

METHODS

5-year follow-up data from 334 patients treated for colorectal cancer were used to train and validate six neural networks designed for the prediction of death within 9, 12, 15, 18, 21, and 24 months. The previously trained 12-month neural network was then applied to 2-year follow-up data from patients from a second institution; outcome was concealed. No further training of the neural network was undertaken. The network's predictions were compared with those of two consultant colorectal surgeons supplied with the same data.

FINDINGS

All six neural networks were able to achieve overall accuracy greater than 80% for the prediction of death for individual patients at institution 1 within 9, 12, 15, 18, 21, and 24 months. The mean sensitivity and specificity were 60% and 88%. When the neural network trained to predict death within 12 months was applied to data from the second institution, overall accuracy of 90% (95% CI 84-96) was achieved, compared with the overall accuracy of the colorectal surgeons of 79% (71-87) and 75% (66-84).

INTERPRETATION

The neural networks were able to predict outcome for individual patients with colorectal cancer much more accurately than the currently available clinicopathological methods. Once trained on data from one institution, the neural networks were able to predict outcome for patients from an unrelated institution.

Neural networks in clinical medicine

Neural networks are parallel, distributed, adaptive information-processing systems that develop their functionality in response to exposure to information. This paper is a tutorial for researchers intending to use neural nets for medical decision-making applications. It includes detailed discussion of the issues particularly relevant to medical data as well as wider issues relevant to any neural net application. The article is restricted to back-propagation learning in multilayer perceptrons, as this is the neural net model most widely used in medical applications.

A predictive index of axillary nodal involvement in operable breast cancer

We investigated the association between pathological characteristics of primary breast cancer and degree of axillary nodal involvement and obtained a predictive index of the latter from the former. In 2076 cases, 17 histological features, including primary tumour and local invasion variables, were recorded. The whole sample was randomly split in a training (75% of cases) and a test sample. Simple and multiple correspondence analysis were used to select the variables to enter in a multinomial logit model to build an index predictive of the degree of nodal involvement. The response variable was axillary nodal status coded in four classes (N0, N1-3, N4-9, N > or = 10). The predictive index was then evaluated by testing goodness-of-fit and classification accuracy. Covariates significantly associated with nodal status were tumour size (P < 0.0001), tumour type (P < 0.0001), type of border (P = 0.048), multicentricity (P = 0.003), invasion of lymphatic and blood vessels (P < 0.0001) and nipple invasion (P = 0.006). Goodness-of-fit was validated by high concordance between observed and expected number of cases in each decile of predicted probability in both training and test samples. Classification accuracy analysis showed that true node-positive cases were well recognised (84.5%), but there was no clear distinction among the classes of node-positive cases. However, 10 year survival analysis showed a superimposible prognostic behaviour between predicted and observed nodal classes. Moreover, misclassified node-negative patients (i.e. those who are predicted positive) showed an outcome closer to patients with 1-3 metastatic nodes than to node-negative ones. In conclusion, the index cannot completely substitute for axillary node information, but it is a predictor of prognosis as accurate as nodal involvement and identifies a subgroup of node-negative patients with unfavourable prognosis.

Artificial neural networks for decision support in clinical medicine

Connectionist models such as neural networks are alternatives to linear, parametric statistical methods. Neural networks are computer-based pattern recognition methods with loose similarities with the nervous system. Individual variables of the network, usually called 'neurones', can receive inhibitory and excitatory inputs from other neurones. The networks can define relationships among input data that are not apparent when using other approaches, and they can use these relationships to improve accuracy. Thus, neural nets have substantial power to recognize patterns even in complex datasets. Neural network methodology has outperformed classical statistical methods in cases where the input variables are interrelated. Because clinical measurements usually derive from multiple interrelated systems it is evident that neural networks might be more accurate than classical methods in multivariate analysis of clinical data. This paper reviews the use of neural networks in medical decision support. A short introduction to the basics of neural networks is given, and some practical issues in applying the networks are highlighted. The current use of neural networks in image analysis, signal processing and laboratory medicine is reviewed. It is concluded that neural networks have an important role in image analysis and in signal processing. However, further studies are needed to determine the value of neural networks in the analysis of laboratory data.