Background: Electronic health records (EHRs) remain a vital source of clinical information, yet processing these heterogeneous data is extremely labor-intensive. Summarization of these data using Large Language Models (LLMs) is considered a promising tool to support practicing physicians. Unbiased, automated quality control is crucial for integrating the tools into routine practice, saving time and labor. This pilot study aimed to assess the potential and constraints of self-contained evaluation of summarization quality (without expert involvement) based on automatic evaluation metrics and LLM-as-a-judge. Methods: The summaries of text data from 30 EHRs were generated by six open-source low-parameter LLMs. The medical summaries were evaluated using standard automatic metrics (BLEU, ROUGE, METEOR, BERTScore) as well as the LLM-as-a-judge approach using the following criteria: relevance, completeness, redundancy, coherence and structure, grammar and terminology, and hallucinations. Expert evaluation was conducted using the same criteria. Results: The results showed that LLMs hold great promise for summarizing medical data. Nevertheless, neither the evaluation metrics nor LLM judges are reliable in detecting factual errors and semantic distortions (hallucinations). In terms of relevance, the Pearson correlation between the summary quality score and the expert opinions was 0.688. Conclusions: Completely automating the evaluation of medical summaries remains challenging. Further research should focus on dedicated methods for detecting hallucinations, along with investigating larger or specialized models trained on medical texts. Additionally, the potential integration of retrieval-augmented generation (RAG) within the LLM-as-a-judge architecture deserves attention. Nevertheless, even now, the combination of LLMs and the automatic evaluation metrics can underpin medical decision support systems by performing initial evaluations and highlighting potential shortcomings for expert review.

Purpose: The aim of this study was evaluating changes in the photosensitizer fluorescence in vivo in the radiation-induced damage area in comparison of intact areas with a simultaneous assessment of changes both in blood parameters and in histological data. Materials and methods: The study was conducted in white outbred SHK mice (n = 21). Their right hindlimbs were irradiated with a dose of 25 Gy after the intraperitoneal injections of photosensitizer ‘Photosens’. Fluorescence intensity was traced in vivo by a laser diagnostic system for seven weeks. Simultaneously, histological examination of the damaged areas and blood tests were performed. Results: An increased intensity of the laser-induced fluorescence of the photosensitizer ‘Photosens’ in the damaged areas, compared to the intact symmetrical ones was observed. Laboratory blood tests and histological examination showed changes that may indicate the occurrence of inflammation. Conclusion: Enhanced intensity of the exogenous fluorescence of the photosensitizer in the radiation-induced inflammation of noncancerous tissues was observed. The obtained results may potentially affect an interpretation of the results of intraoperative tumors navigation that have been previously irradiated and can be used for selection of an individualized dose fractionation algorithm in radiology.

According to modern ideas, a reasonable choice of an effective method of treating plaque scleroderma is based on the diagnosis of the pathological process prevailing in the tissues (inflammation-sclerosis). Therefore, an urgent problem of a personalized approach to dermatosis therapy is the possibility of an objective assessment of the prevailing process using non-invasive diagnostic methods. The article presents a clinical case of widespread plaque scleroderma in a 66-year-old patient, demonstrating the possibility of using laser fluorescence spectroscopy and laser Doppler flowmetry to determine the degree of activity of the focus and determine the leading pathological process. We selected three pathological skin foci localized in the abdomen and characterizing three clinical stages of the disease (inflammation, induration, sclerosis). The analysis of fluorescence and laser Doppler flowmetry data showed that in areas clinically defined as inflammation, there is an increase in the average values of the indices of tissue content of porphyrins, lipofuscin and microcirculation index compared with intact skin, while the intensity of collagen fluorescence does not differ significantly. In the induration zone, along with an increase in the fluorescence indices of lipofuscin and porphyrins, there is an increase in the average values of collagen fluorescence indices at effective registration waves. The data obtained by us indicate an active inflammatory process in these foci and the process of fibrosis in the induration zone. In the sclerosis zone, there is an increase in the average values of collagen fluorescence indices compared with intact skin, and the fluorescence of optical markers of inflammation (lipofuscin and porphyrins) do not differ significantly in comparison with the control intact skin. When analyzing the fluorescence spectra and laser Doppler flowmetry data after treatment, we found that in the zones of induration and inflammation, the average values of the fluorescence indices of porphyrins, lipofuscin, collagen and microcirculation index are reduced relative to the initial values (before treatment), but remain higher in comparison with intact skin. The data obtained may indicate that active inflammation in these foci persists at the time of the study. In the study of the focus of sclerosis, the obtained autofluorescence and microcirculation data in dynamics do not differ significantly from the initial values. The data for laser Doppler flowmetry and laser fluorescence spectroscopy are consistent with ultrasound examination of the skin. In our study, the potential possibility of using laser fluorescence spectroscopy and laser Doppler flowmetry methods to establish the degree of activity of the focus, determine the leading pathological process, as well as to evaluate the effectiveness of therapy was demonstrated for the first time.

Keloid scars, in contrast to other scar types, significantly reduce the patient's quality of life. To develop a nondestructive optical diagnostic technique predicting the keloid scars formation in vivo, laser-induced fluorescence spectroscopy (LFS) was used to study the autofluorescence in skin of patients with various types of head and neck cicatricial deformities. The unexpected results were obtained for the endogenous fluorescence of lipofuscin. Significantly reduced autofluorescence of lipofuscin was registered both in the intact and in the keloid scar tissues in comparison with the intact and scar tissues in patients with hypertrophic and normotrophic scars. Sensitivity and specificity achieved by LFS in keloid diagnosis are 81.8% and 93.9% respectively. It could take place due to the changes in the reductive-oxidative balance in cells, as well as due to the proteolysis processes violation. Therefore, we suppose that the evaluation of the lipofuscin autofluorescence in skin before any surgical intervention could predict the probability of the subsequent keloid scars formation.

The paper considers a problem of determination of the luminophore relative concentration in a turbid medium using a laser fluorescence spectroscopy and a diffuse reflectance spectroscopy. The proposed approach is based on the improved Kubelka-Munk theoretical model with the use of experimental data obtained from turbid biological phantom media. Influence of scattering on the radiation registered by a detector is also investigated providing calibration curves to solve the problem.

BACKGROUND: Chest X-ray is a widely used and accessible screening method for lung pathology. However, its diagnostic capabilities, particularly for visualizing pulmonary nodules, are limited. Suboptimal detection of pulmonary nodules by radiologists on chest X-rays affects the timeliness of diagnosis and therapeutic outcomes. One approach to improving lung nodule detection in chest X-rays involves adopting novel techniques, particularly those based on artificial intelligence. Nevertheless, the diagnostic value of these technologies in clinical practice remains largely unresolved due to limited data on radiologists’ performance conventional metrics. AIM: To evaluate the diagnostic performance of 100 radiologists in identifying pulmonary nodules on chest X-rays. METHODS: Each of the 100 radiologists was asked to evaluate 100 chest radiographs, of which 50 showed abnormal findings and 50 were normal. The presence of pulmonary nodules was assessed using the following scale: Absent (0 on the probability scale), likely absent (0.25), uncertain (0.50), likely present (0.75), present (1.00). Validation of nodule presence or absence was performed by three expert physicians using a binary scale (0/1) based on chest CT performed within 14 days after chest X-ray. We assessed image interpretation time, the difference in performance between radiologists and expert physicians (expressed in absolute units as D), and the primary diagnostic accuracy metrics of the radiologists. RESULTS: The study yielded the following: ROC AUC: 0.858 ± 0.059, accuracy: 0.822 ± 0.048, sensitivity: 0.779 ± 0.097, and specificity: 0.864 ± 0.095. A negligible positive correlation was observed between expert accuracy and mean image interpretation time (Spearman correlation coefficient rs = 0.189) and a low positive correlation (rs = 0.344) between image interpretation time and the D value. CONCLUSION: These results can be used to assess the quality of automated detection systems under development and to evaluate the efficacy of alternative methods and approaches for pulmonary nodule detection.