Interpretable Machine Learning in Healthcare

Ahmad, Muhammad Aurangzeb; Teredesai, Ankur; Eckert, Carly

doi:10.1109/ichi.2018.00095

Cited by 109 publications

(87 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Clinical machine learning systems must be interpretable by human users. 28 Many approaches for improving interpretability among "black box" systems have been developed in the machine learning literature; for this study, we opted to use the Gradient Class Activation Maps (Grad-CAM) algorithm, which identifies parts of an input which were important in the model's final decision about that input. 29 We used Grad-CAM to generate saliency maps for our model's predictions to visualize the importance of each word token in the model's decision making process.…”

Section: Model Interpretabilitymentioning

confidence: 99%

A Web Application for Adrenal Incidentaloma Identification, Tracking, and Management Using Machine Learning

et al. 2020

View full text Add to dashboard Cite

Background Incidental radiographic findings, such as adrenal nodules, are commonly identified in imaging studies and documented in radiology reports. However, patients with such findings frequently do not receive appropriate follow-up, partially due to the lack of tools for the management of such findings and the time required to maintain up-to-date lists. Natural language processing (NLP) is capable of extracting information from free-text clinical documents and could provide the basis for software solutions that do not require changes to clinical workflows. Objectives In this manuscript we present (1) a machine learning algorithm we trained to identify radiology reports documenting the presence of a newly discovered adrenal incidentaloma, and (2) the web application and results database we developed to manage these clinical findings. Methods We manually annotated a training corpus of 4,090 radiology reports from across our institution with a binary label indicating whether or not a report contains a newly discovered adrenal incidentaloma. We trained a convolutional neural network to perform this text classification task. Over the NLP backbone we built a web application that allows users to coordinate clinical management of adrenal incidentalomas in real time. Results The annotated dataset included 404 positive (9.9%) and 3,686 (90.1%) negative reports. Our model achieved a sensitivity of 92.9% (95% confidence interval: 80.9–97.5%), a positive predictive value of 83.0% (69.9–91.1)%, a specificity of 97.8% (95.8–98.9)%, and an F1 score of 87.6%. We developed a front-end web application based on the model's output. Conclusion Developing an NLP-enabled custom web application for tracking and management of high-risk adrenal incidentalomas is feasible in a resource constrained, safety net hospital. Such applications can be used by an institution's quality department or its primary care providers and can easily be generalized to other types of clinical findings.

show abstract

Section: Model Interpretabilitymentioning

confidence: 99%

A Web Application for Adrenal Incidentaloma Identification, Tracking, and Management Using Machine Learning

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Indeed, a salient one is the degree to which a human being (e.g., a physician) can really understand the actual cause of a decision made by an AI system (e.g., the diagnostic of a disease). Such attribute is called interpretability [1][2][3].…”

Section: Introductionmentioning

confidence: 99%

Property Checking with Interpretable Error Characterization for Recurrent Neural Networks

Mayr

Yovine

Visca

2021

MAKE

View full text Add to dashboard Cite

This paper presents a novel on-the-fly, black-box, property-checking through learning approach as a means for verifying requirements of recurrent neural networks (RNN) in the context of sequence classification. Our technique steps on a tool for learning probably approximately correct (PAC) deterministic finite automata (DFA). The sequence classifier inside the black-box consists of a Boolean combination of several components, including the RNN under analysis together with requirements to be checked, possibly modeled as RNN themselves. On one hand, if the output of the algorithm is an empty DFA, there is a proven upper bound (as a function of the algorithm parameters) on the probability of the language of the black-box to be nonempty. This implies the property probably holds on the RNN with probabilistic guarantees. On the other, if the DFA is nonempty, it is certain that the language of the black-box is nonempty. This entails the RNN does not satisfy the requirement for sure. In this case, the output automaton serves as an explicit and interpretable characterization of the error. Our approach does not rely on a specific property specification formalism and is capable of handling nonregular languages as well. Besides, it neither explicitly builds individual representations of any of the components of the black-box nor resorts to any external decision procedure for verification. This paper also improves previous theoretical results regarding the probabilistic guarantees of the underlying learning algorithm.

show abstract

“…However, the rapid growth DATA ANALYSIS AND INTELLIGENCE SYSTEMS of available data and digital images revealed the need for automatic determination of the patient's clinical pathway based on these data. The solution for the automatic identification of a personal clinical pathway was the technology of process mining [7,8], data mining [9], machine learning algorithms [10,11] and others.…”

Section: Introductionmentioning

confidence: 99%

Clinical pathways analysis of patients in medical institutions based on hard and fuzzy clustering methods

Prokofyeva¹,

Zaytsev²

2020

Bus. Inform.

View full text Add to dashboard Cite

Modeling the processes in a healthcare system plays a large role in understanding its activities and serves as the basis for increasing the efficiency of medical institutions. The tasks of analyzing and modeling large amounts of urban healthcare data using machine learning methods are of particular importance and relevance for the development of industry solutions in the framework of digitalization of the economy, where data is the key factor in production. The problem of automatic analysis and determination of clinical pathways groups of patients based on clustering methods is considered in this research. Existing projects in this area reflect a great interest on the part of the scientific community in such studies; however, there is a need to develop a number of methodological approaches for their further practical application in urban outpatient institutions, taking into account the specifics of the organization being analyzed. The aim of the study is to improve the quality of management and segmentation of patient input flow in urban medical institutions based on cluster analysis methods for the further development of recommendation services. One approach to achieving this goal is the development and implementation of clinical pathways, or patient trajectories. In general, the clinical pathway of a patient might be interpreted as the trajectory when receiving medical services in respective institutions. The approach of developing groups of patient routes by the hierarchical agglomerative algorithm with the Ward method and Additive Regularization of Topic Models (ARTM) is presented in this article. A computational experiment based on public data on the routes of patients with a DATA ANALYSIS AND INTELLIGENCE SYSTEMS

show abstract

Interpretable Machine Learning in Healthcare

Cited by 109 publications

References 0 publications

A Web Application for Adrenal Incidentaloma Identification, Tracking, and Management Using Machine Learning

A Web Application for Adrenal Incidentaloma Identification, Tracking, and Management Using Machine Learning

Property Checking with Interpretable Error Characterization for Recurrent Neural Networks

Clinical pathways analysis of patients in medical institutions based on hard and fuzzy clustering methods

Contact Info

Product

Resources

About