This article systematically reviewed AutoML techniques, summarized interpretation methods, and detailed how AutoML with interpretation has been used for six major healthcare data types: image, free text, tabular data, signal, genomic sequences, and multi-modality.

Our work revealed that molecular subtypes based on nuclear mitochondria-related genes (MTRGs) in Clear cell renal cell carcinoma (ccRCC) were related to patient prognosis and immune infiltration levels. Moreover, a critical nuclear MTRG, succinate-CoA ligase ADP-forming subunit beta (SUCLA2), was identified, which potentially represented an important risk factor in ccRCC. In brief, this study provided new insights and guidance for the clinical diagnosis of ccRCC.

This study identifies a 7 mL volume of agitated saline as an optimal balance, maintaining diagnostic accuracy for Patent foramen ovale detection while reducing the physical strain on operators during the procedure.

This study forecasts the transmissibility of specific amino acid mutations that have emerged in the Omicron variant of SARS-CoV-2. Through predictive modeling, we identify mutations with high transmission potential. Subsequently, we assess these mutations for conservation and their impact on protein structure and function, isolating those that pose significant risk. Based on the intrinsic properties of these high-risk mutations, we delineate the underlying drivers that have facilitated the global spread of the Omicron variants.

This study investigates the escalated challenges confronting patient navigators following the COVID-19 pandemic and assesses the potential of AI-enhanced tools to alleviate these issues. It uncovers significant gaps in technological support for navigators, who are essential in managing diverse barriers faced by patients. The findings highlight the potential of tailored AI interventions that could significantly boost the capabilities and effectiveness of patient navigators, thereby supporting the long-term sustainability of this service and overall healthcare systems.

This article provides a comparison between large language models differing in their online web search capability and their performance on the Orthopaedic In-Training Examination. Bing AI GPT 4.0 scored higher compared to previous ChatGPT 3.5, and higher than the average performance of orthopaedic residents in national accredited programs. Generative AI can provide a logical context across answer responses through its in-depth information searches and citation of resources.

Advanced T stage, submandibular gland location, and unpaired lesions were associated with lymph node involvement. The prognosis of SDC patients with PLNR ≤ 0.48 was better than that of SDC patients with PLNR > 0.48. PLNR can reflect both the effect of lymph node dissection and the number of positive lymph nodes, and its prognostic value for major SDC exceeds the number of positive lymph node number. Patients with high PLNR should be followed closely after surgery.

This prospective study showed that updated knowledge regarding the management modalities of PAC is a prerequisite for accurate histopathological assessment and good surgical outcomes. Furthermore, histopathological analysis helps clinicians to select effective postoperative therapies for greater improvement in survival outcomes.

A 31-year-old woman with anti-synthetase syndrome developed abdominal and subcutaneous lumps during immunosuppressive treatment, with elevated tumor markers suggesting malignancy. Pathological examination revealed Mycobacterium tuberculosis, leading to a revised diagnosis of multifocal extrapulmonary tuberculosis. Targeted therapy improved her condition. This case underscores the necessity of thorough pathological evaluations and broad differential diagnosis in patients with overlapping infectious and autoimmune symptoms.

This article performs a literature search to determine the current progress, identifies research gaps, and highlights future opportunities of human-in-the-loop across the machine learning lifecycle, including data preparation, feature engineering, model development, and model deployment.