Artificial intelligence for prediction of treatment outcomes in breast cancer: Systematic review of design, reporting standards, and bias.

BACKGROUND: Artificial intelligence (AI) has the potential to personalize treatment strategies for patients with cancer. However, current methodological weaknesses could limit clinical impact. We identified common limitations and suggested potential solutions to facilitate translation of AI to breast cancer management. METHODS: A systematic review was conducted in MEDLINE, Embase, SCOPUS, Google Scholar and PubMed Central in July 2021. Studies investigating the performance of AI to predict outcomes among patients undergoing treatment for breast cancer were included. Algorithm design and adherence to reporting standards were assessed following the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement. Risk of bias was assessed by using the Prediction model Risk Of Bias Assessment Tool (PROBAST), and correspondence with authors to assess data and code availability. RESULTS: Our search identified 1,124 studies, of which 64 were included: 58 had a retrospective study design, with 6 studies with a prospective design. Access to datasets and code was severely limited (unavailable in 77% and 88% of studies, respectively). On request, data and code were made available in 28% and 18% of cases, respectively. Ethnicity was often under-reported (not reported in 52 of 64, 81%), as was model calibration (63/64, 99%). The risk of bias was high in 72% (46/64) of the studies, especially because of analysis bias. CONCLUSION: Development of AI algorithms should involve external and prospective validation, with improved code and data availability to enhance reliability and translation of this promising approach. Protocol registration number: PROSPERO - CRD42022292495.

Identification of Characteristic Points in Multivariate Physiological Signals by Sensor Fusion and Multi-Task Deep Networks.

Identification of characteristic points in physiological signals, such as the peak of the R wave in the electrocardiogram and the peak of the systolic wave of the photopletismogram, is a fundamental step for the quantification of clinical parameters, such as the pulse transit time. In this work, we presented a novel neural architecture, called eMTUnet, to automate point identification in multivariate signals acquired with a chest-worn device. The eMTUnet consists of a single deep network capable of performing three tasks simultaneously: (i) localization in time of characteristic points (labeling task), (ii) evaluation of the quality of signals (classification task); (iii) estimation of the reliability of classification (reliability task). Preliminary results in overnight monitoring showcased the ability to detect characteristic points in the four signals with a recall index of about 1.00, 0.90, 0.90, and 0.80, respectively. The accuracy of the signal quality classification was about 0.90, on average over four different classes. The average confidence of the correctly classified signals, against the misclassifications, was 0.93 vs. 0.52, proving the worthiness of the confidence index, which may better qualify the point identification. From the achieved outcomes, we point out that high-quality segmentation and classification are both ensured, which brings the use of a multi-modal framework, composed of wearable sensors and artificial intelligence, incrementally closer to clinical translation.