Feldt, Robert

AbstractArtificial intelligence (AI) has been influencing healthcare and medical research for several years and will likely become indispensable in the near future. AI is intended to support healthcare professionals to make the healthcare system more efficient and ultimately improve patient outcomes. Despite the numerous benefits of AI systems, significant concerns remain. Errors in AI systems can pose serious risks to human health, underscoring the critical need for safety, as well as adherence to ethical and moral standards, before these technologies can be integrated into clinical practice. To address these challenges, the development, certification, and deployment of medical AI systems must adhere to strict and transparent regulations. The European Commission has already established a regulatory framework for AI systems by enacting the European Union Artificial Intelligence Act. This review article, part of an AI learning series, discusses key considerations for medical AI systems such as reliability, accuracy, trustworthiness, lawfulness and legal compliance, ethical and moral alignment, sustainability, and regulatory oversight.Level of EvidenceLevel V.

Abstract Multimodal artificial intelligence (AI) has the potential to revolutionise healthcare by enabling the simultaneous processing and integration of various data types, including medical imaging, electronic health records, genomic information and real‐time data. This review explores the current applications and future potential of multimodal AI across healthcare, with a particular focus on orthopaedic surgery. In presurgical planning, multimodal AI has demonstrated significant improvements in diagnostic accuracy and risk prediction, with studies reporting an Area under the receiving operator curve presenting good to excellent performance across various orthopaedic conditions. Intraoperative applications leverage advanced imaging and tracking technologies to enhance surgical precision, while postoperative care has been advanced through continuous patient monitoring and early detection of complications. Despite these advances, significant challenges remain in data integration, standardisation, and privacy protection. Technical solutions such as federated learning (allowing decentralisation of models) and edge computing (allowing data analysis to happen on site or closer to site instead of multipurpose datacenters) are being developed to address these concerns while maintaining compliance with regulatory frameworks. As this field continues to evolve, the integration of multimodal AI promises to advance personalised medicine, improve patient outcomes, and transform healthcare delivery through more comprehensive and nuanced analysis of patient data. Level of Evidence: Level V.

Abstract Explorative data analysis (EDA) is a critical step in scientific projects, aiming to uncover valuable insights and patterns within data. Traditionally, EDA involves manual inspection, visualization, and various statistical methods. The advent of artificial intelligence (AI) and machine learning (ML) has the potential to improve EDA, offering more sophisticated approaches that enhance its efficacy. This review explores how AI and ML algorithms can improve feature engineering and selection during EDA, leading to more robust predictive models and data‐driven decisions. Tree‐based models, regularized regression, and clustering algorithms were identified as key techniques. These methods automate feature importance ranking, handle complex interactions, perform feature selection, reveal hidden groupings, and detect anomalies. Real‐world applications include risk prediction in total hip arthroplasty and subgroup identification in scoliosis patients. Recent advances in explainable AI and EDA automation show potential for further improvement. The integration of AI and ML into EDA accelerates tasks and uncovers sophisticated insights. However, effective utilization requires a deep understanding of the algorithms, their assumptions, and limitations, along with domain knowledge for proper interpretation. As data continues to grow, AI will play an increasingly pivotal role in EDA when combined with human expertise, driving more informed, data‐driven decision‐making across various scientific domains. Level of Evidence: Level V ‐ Expert opinion.

AbstractArtificial intelligence's (AI) accelerating progress demands rigorous evaluation standards to ensure safe, effective integration into healthcare's high‐stakes decisions. As AI increasingly enables prediction, analysis and judgement capabilities relevant to medicine, proper evaluation and interpretation are indispensable. Erroneous AI could endanger patients; thus, developing, validating and deploying medical AI demands adhering to strict, transparent standards centred on safety, ethics and responsible oversight. Core considerations include assessing performance on diverse real‐world data, collaborating with domain experts, confirming model reliability and limitations, and advancing interpretability. Thoughtful selection of evaluation metrics suited to the clinical context along with testing on diverse data sets representing different populations improves generalisability. Partnering software engineers, data scientists and medical practitioners ground assessment in real needs. Journals must uphold reporting standards matching AI's societal impacts. With rigorous, holistic evaluation frameworks, AI can progress towards expanding healthcare access and quality.Level of EvidenceLevel V.

The purpose of this study was to re-affirm our clinical impression that non-institutionalized children whose head circumference was below minus 2 standard deviations were mentally subnormal and frequently had growth failure. A group of 134 children with a head circumference below minus 2 standard deviations from the mean were studied, and all but one were mentally subnormal. The most severe mental retardation was noted in the group of children with a head circumference of minus 4 standard deviations or below. We found, as have others, that children with mental retardation have height and weights below the expected norm and that children with a head circumference below minus 2 standard deviations have even lower mean heights and weights. The head circumference of 31 children with growth failure and normal intelligence was normal for age and sex, therefore disproving the concept that the abnormally small child has a proportionally small head. In the child with growth failure, should the head be proportionally small (below minus 2 standard deviations), mental subnormality should be suspected. We feel that the head circumference measurement has taken on new clinical significance in that our data support its use in suspecting the association of mental subnormality in children with growth failure and a head circumference of below minus 2 standard deviations from the mean for age and sex.

Replying to Dr. I. Krieger's letter to the editor poses difficulties. In using this communication for submission of her own data for publication, she has not felt the need to provide the careful documentation that would allow critical analysis. For example, her statement that "small stature, small head size, and mental subnormality are frequently associated" is true on the basis of our experience but is not supported by her recorded data. No psychometric test results are noted in the text or in the illustrations accompayning her letter.

In right-sided extracardiac conduits inserted for surgical correction of certain congenital heart defects, obstruction is a significant late complication requiring conduit replacement in about 5% of the patients. Our histopathologic study indicates a significant role of platelets in the pathogenesis. This study was designed to evaluate the feasibility of imaging procedure with 111-Indium-labeled platelets to localize and later quantify platelet accumulation in the conduits. Four patients with pulmonary atresia and 1 with truncus arteriosus underwent conduit placement, 2 with and 3 without porcine valve. Autologous platelets collected.from 43 ml of blood drawn prior to surgery were labeled with 111-Indium-oxine in ACD-saline medium (labeling efficiency 50-60%). 400-500 μCi of autologous 111-Indium-labeled platelets were administered intravenously immediately after surgery in 4, and five days after surgery in 1 patient. Imaging was performed at 6 hours and daily for 1-4 days using a portable gamma camera interfaced with a computer and fitted with medium-energy parallel hole collimator. Platelet deposition was noted at the site of the conduit in all 5 patients. Maximum accumulation (0.1-0.3% of dose) of platelets was seen within the first 24 hours. Activity over blood pool and lungs declined and over liver and spleen increased in a predictable manner over the imaging period. This technique has potential use as a noninvasive quantitative tool to study the effect of platelet inhibitor drugs in preventing conduit obstruction.

Serial histologic sections of the atrioventricular (A-V) conduction system were studied in four hearts with persistent common A-V canal, and the anatomic findings were related to the abnormal ECG patterns. Two hearts were examples of the partial form and two of the complete form of A-V canal. The major anatomic abnormalities were: (1) postero-inferior displacement of the A-V node; (2) relatively short distances between the A-V node and the origin of the left bundle branching; (3) marked posteroinferior displacement of the left bundle branching system; and (4) relative hypoplasia of the anterior (superior) left bundle branches. The anatomic configuration of the A-V conduction system in A-V canal could result in relatively early impulse propagation to the posterior aspect of the ventricular septum as suggested by reported electrophysiologica studies. The anatomic findings are considered to be related closely to the known ECG patterns of A-V canal.

The atrioventricular conduction system was studied by a technique of selected serial histological sections in examples of tetralogy of Fallot and ventricular septal defect to correlate the anatomic findings observed with the electrophysiological findings as manifested by the vectorcardiographic pattern. Eight hearts were studied; two examples of each anomaly had electrocardiographic and VCG patterns in the frontal plane similar to A-V canal defects and two examples of each had ECG and VCG patterns usually associated with these lesions. Of the four examples of tetralogy, the two examples with VCG patterns similar to the A-V canal pattern differed from the other examples in that the distance between the A-V node and the origin of left bundle branching was shorter and the fibers destined to become the right bundle branch took an elongated course. The VSD examples offered less striking differences. A short distance from the A-V node to the onset of left bundle branching was seen in both examples with VCG patterns similar to the A-V canal pattern, but one example of VSD with a representative VCG pattern had a similar anatomic configuration. When measurement ratios were compared, the right bundle took its origin at a relatively greater distance from the A-V node in the examples with VCG patterns similar to the A-V canal pattern. The anatomic configuration of those examples of tetralogy and VSD with VCG patterns similar to the A-V canal pattern was such that relatively early conduction to the left bundle branch system could occur.

Records of 273 cases of ventricular septal defect from five participating institutions were studied. Insignificant hemodynamic disturbance (Roger's) was found in 11 per cent, moderate to large left-to-right shunts in 55 per cent, and severe pulmonary vascular disease ("Eisenmenger's") in 8 per cent. Because of insufficient data, 26 per cent of the cases were not classified hemodynamically. Thus, a spectrum of severity ranging from mild to severe was noted. There was evidence of congestive heart failure in infancy in 78 cases and a significant mortality occurred in this group, whether treatment was medical (mortality, 27 per cent) or surgical (mortality, 46 per cent). Six patients had hemodynamic evidence of severe progressive pulmonary vascular disease, three when they were less than 8 years old. Correlations of hemodynamic severity with clinical symptoms, roentgenographic findings, and electrocardiographic changes were reasonably good at the extreme ends of the severity spectrum but only fair in the middle range. Over-all mortality was high (28 per cent) for the 163 patients treated surgically, but mortality was also high (18 per cent) for the 110 patients treated medically. Ten of the 20 deaths in the medical group occurred in patients who had had congestive heart failure in infancy. Further study of the natural history of ventricular septal defect is clearly indicated for establishment of improved criteria for selection of cases for operation.