Climate change and artificial intelligence in healthcare: Review and recommendations towards a sustainable future.

The rapid advancement of artificial intelligence (AI) in healthcare has revolutionized the industry, offering significant improvements in diagnostic accuracy, efficiency, and patient outcomes. However, the increasing adoption of AI systems also raises concerns about their environmental impact, particularly in the context of climate change. This review explores the intersection of climate change and AI in healthcare, examining the challenges posed by the energy consumption and carbon footprint of AI systems, as well as the potential solutions to mitigate their environmental impact. The review highlights the energy-intensive nature of AI model training and deployment, the contribution of data centers to greenhouse gas emissions, and the generation of electronic waste. To address these challenges, the development of energy-efficient AI models, the adoption of green computing practices, and the integration of renewable energy sources are discussed as potential solutions. The review also emphasizes the role of AI in optimizing healthcare workflows, reducing resource waste, and facilitating sustainable practices such as telemedicine. Furthermore, the importance of policy and governance frameworks, global initiatives, and collaborative efforts in promoting sustainable AI practices in healthcare is explored. The review concludes by outlining best practices for sustainable AI deployment, including eco-design, lifecycle assessment, responsible data management, and continuous monitoring and improvement. As the healthcare industry continues to embrace AI technologies, prioritizing sustainability and environmental responsibility is crucial to ensure that the benefits of AI are realized while actively contributing to the preservation of our planet.

Maximum standardized uptake value in 11C-methionine positron emission tomography may predict the prognosis of patients with oral squamous cell carcinoma.

The present study aimed to elucidate the correlation between the uptake of 11C-methionine (MET) by a primary tumor and the survival of patients with oral squamous cell carcinoma (OSCC). This study enrolled 31 patients who underwent radical surgery for OSCC. The patients underwent pretreatment MET-positron emission tomography (PET) scanning. We analyzed correlations between the maximum standardized uptake value (SUVmax) of MET-PET in a primary tumor and the clinicopathological features. Further, we compared overall survival (OS), disease-specific survival (DSS), and loco-regional recurrence (LRR) rates between the two groups according to SUVmax of MET-PET. SUVmax of MET-PET in a primary tumor was higher in patients with advanced T-classification and advanced clinical stage, with significant differences (P = 0.001 and P = 0.016, respectively). The patients with SUVmax of MET-PET ≥ 4.4 showed significantly lower DSS rates and higher LRR rates than those with SUVmax of < 4.4 (P = 0.015 and P = 0.016, respectively). SUVmax of MET-PET and OS rates showed no significant correlation (P = 0.073). The present study revealed that SUVmax of MET-PET may predict clinical outcomes and prognosis in patients with OSCC who underwent radical surgery.

Safety and efficacy of multiple-dose versus single-dose MIBG therapy in patients with refractory pheochromocytoma and paraganglioma: a single-center retrospective analysis.

OBJECTIVE: To investigate the incidence of adverse events (AEs) following single and multiple administrations of I-131 metaiodobenzylguanidine (MIBG) therapy for inoperable pheochromocytomas and paragangliomas (PPGLs). METHODS: A single-center retrospective study was conducted on patients with inoperable PPGLs who underwent I-131 MIBG therapy between January 2000 and December 2020. A total of 28 patients with available electronic medical records were included. The treatment consisted of a single intravenous administration of 150 mCi (5.55 GBq) of I-131 MIBG. We evaluated the first MIBG treatment and repeated MIBG treatments performed within 200 days of the previous treatment. AEs for each treatment were evaluated using CTCAE version 4.0, and the statistical analysis was conducted at a significance level of p < 0.05. Objective response based on RECIST 1.1 criteria and biochemical response based on urinary catecholamines were assessed. RESULTS: The study included a total of 63 administrations, consisting of 28 single administrations (SAs), including the first administration for all 28 cases, and 35 multiple administrations (MAs), which included the second or later administrations. Hematological AEs were evaluable for 23 SAs and 29 MAs. Grade 3 or higher leukopenia occurred in 9.8% of all administrations, and Grade 3 or higher lymphopenia in 23.5%; both were manageable through observation. There were no significant differences in clinical AE Grades 1-2 (p = 0.32), hematological AE Grades 1-2 (p = 0.22), or hematological AE Grades 3-4 (p = 0.12) between MAs and SAs. Statistical analysis for each type of AE revealed significant increases in leukopenia (p < 0.01) and lymphopenia (p = 0.04). No significant difference in anemia, thrombocytopenia, or neutropenia was observed between MAs and SAs. There was no significant increase in the incidence rate of Grade 3 or higher hematological AEs for any of the parameters. The objective response rate was 0% for SAs and 36% for MAs. Biochemical response rates were 18% for SAs and 67% for MAs. CONCLUSION: In I-131 MIBG therapy for PPGLs, multiple administrations significantly increased only Grade 1 or 2 lymphopenia and leukopenia compared to single administration.

Improvement in the estimation of perfusable tissue fraction and myocardial flow reserve in the ischemic myocardial lesions using ECG-gated dynamic myocardial PET with 15O-water.

OBJECTIVE: Perfusable tissue fraction (PTF) and myocardial flow reserve (MFR) from 15O-water dynamic positron emission tomography (PET) are parameters of myocardial viability. However, myocardial motion causes errors in these values. We aimed to develop accurate estimation of PTF and MFR in ischemic lesions using an electro-cardiogram (ECG)-gated dynamic myocardial PET with 15O-water. METHODS: Twenty-seven patients with ischemic heart disease were enrolled. All patients underwent percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG). List mode 3D PET data and ECG signals were acquired using Philips Gemini TF64 instrument. For each scan, 500 MBq of 15O-water was infused slowly for 2 min, and the dynamic data were scanned for 6 min. Both non-gated dynamic images and ECG-gated diastolic dynamic images were reconstructed. On the myocardial PET images of each patient, the entire myocardial region of interest (ROI) was set and divided into 17 segments. Myocardial blood flow in the resting state (rest MBF), hyperemic state (stress MBF), PTF, and MFR in each segment were estimated from both non-gated and ECG-gated dynamic PET images. Coronary arteriograms were obtained for all patients. In total, 128 normal segments without stenosis and 50 ischemic segments with > 90% stenosis were evaluated. RESULTS: In the ischemic myocardial segments, the PTF with ECG-gated PET was estimated as significantly lower than that with non-gated PET (0.63 ± 0.09 vs. 0.72 ± 0.08 [mL/mL], p < 0.001). The ECG-gated PET estimated a significantly lower PTF in the ischemic segments than in the normal segments (0.63 ± 0.09 vs. 0.67 ± 0.07 [mL/mL], p < 0.01). In the normal segments, the ECG-gated PET detected no significant difference in MFR compared with those from the non-gated PET (2.15 ± 0.76 vs. 2.24 ± 0.79, p = 0.28). However, in the ischemic myocardial segments, the MFR with ECG-gated PET was estimated as significantly lower than that with the non-gated PET (1.23 ± 0.29 vs. 1.69 ± 0.71, p < 0.001). The ECG-gated PET presented a significantly higher inter-observer reproducibility of PTF and rest MBF than the non-gated PET (p < 0.01). Neither stress MBF nor MFR yielded significant differences in inter-observer reproducibility between the ECG-gated and non-gated PET. CONCLUSIONS: The ECG-gated dynamic 15O-water PET suppressed the myocardial motion effect and resulted in a lower PTF and MFR in ischemic myocardial lesions than the non-gated PET. The ECG-gated PET seemed to be better than the conventional non-gated dynamic PET for the detection of ischemic myocardial lesion.

Stability of standardized uptake values for quantitative bone SPECT for jawbone lesions: a single-center cross-sectional study.

BACKGROUND: The long time required for bone uptake of radiopharmaceutical material after injection for bone scintigraphy is a burden for patients with poor health. Thus, to assess whether the uptake time could be reduced for single-photon emission computed tomography (SPECT) of the jawbone, this study evaluated differences in maximum standardized uptake values (SUVmax) within patients using SPECT imaging at 2 and 3 hours after radiopharmaceutical injection. METHODS: A total of 33 patients undergoing treatment or in post-treatment follow-up for medication-related osteonecrosis of the jaw, who visited our hospital between July 2020 and August 2021 and could receive SPECT twice on the same day, were enrolled in the study. Patients were injected with technetium-99 m hydroxymethylene diphosphonate (Tc-99 m HMDP) intravenously. The SUVmax for healthy parietal bones and jawbone lesions were calculated from the SPECT images using quantitative analysis software, and the SUVmax were compared between 2- and 3-hour uptake times. RESULTS: After exclusion, 30 patients were included in the study. In the 2-hour and 3-hour images, the median SUVmax of the parietal bones were 1.90 and 1.81, respectively, and those of the jawbone lesions were 9.25 and 9.39, respectively. The limits of agreement (LOA) ranged from - 0.33 to 0.25 in the parietal bones, and the %LOA ranged from - 9.8 to 17.3% in the jawbone lesions, showing high equivalence between the two uptake durations. The SUVmax showed no clinical differences between the 2- and 3-hour uptake durations for Tc-99 m HMDP SPECT of the jawbone. CONCLUSIONS: The results of this study justify a 2-3-hour uptake window when performing quantitative SPECT of the jawbone. Therefore, the minimum uptake time can potentially be reduced to only 2 hours.

Identifying G6PC3 as a Potential Key Molecule in Hypoxic Glucose Metabolism of Glioblastoma Derived from the Depiction of 18F-Fluoromisonidazole and 18F-Fluorodeoxyglucose Positron Emission Tomography.

Purpose: Glioblastoma is the most aggressive primary brain tumor, characterized by its distinctive intratumoral hypoxia. Sequential preoperative examinations using fluorine-18-fluoromisonidazole (18F-FMISO) and fluorine-18-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) could depict the degree of glucose metabolism with hypoxic condition. However, molecular mechanism of glucose metabolism under hypoxia in glioblastoma has been unclear. The aim of this study was to identify the key molecules of hypoxic glucose metabolism. Methods: Using surgically obtained specimens, gene expressions associated with glucose metabolism were analyzed in patients with glioblastoma (n = 33) who underwent preoperative 18F-FMISO and 18F-FDG PET to identify affected molecules according to hypoxic condition. Tumor in vivo metabolic activities were semiquantitatively evaluated by lesion-normal tissue ratio (LNR). Protein expression was confirmed by immunofluorescence staining. To evaluate prognostic value, relationship between gene expression and overall survival was explored in another independent nonoverlapping clinical cohort (n = 17) and validated by The Cancer Genome Atlas (TCGA) database (n = 167). Results: Among the genes involving glucose metabolic pathway, mRNA expression of glucose-6-phosphatase 3 (G6PC3) correlated with 18F-FDG LNR (P = 0.03). In addition, G6PC3 mRNA expression in 18F-FMISO high-accumulated glioblastomas was significantly higher than that in 18F-FMISO low-accumulated glioblastomas (P < 0.01). Protein expression of G6PC3 was consistent with mRNA expression, which was confirmed by immunofluorescence analysis. These findings indicated that the G6PC3 expression might be facilitated by hypoxic condition in glioblastomas. Next, we investigated the clinical relevance of G6PC3 in terms of prognosis. Among the glioblastoma patients who received gross total resection, mRNA expressions of G6PC3 in the patients with poor prognosis (less than 1-year survival) were significantly higher than that in the patients who survive more than 3 years. Moreover, high mRNA expression of G6PC3 was associated with poor overall survival in glioblastoma, as validated by TCGA database. Conclusion: G6PC3 was affluently expressed in glioblastoma tissues with coincidentally high 18F-FDG and 18F-FMISO accumulation. Further, it might work as a prognostic biomarker of glioblastoma. Therefore, G6PC3 is a potential key molecule of glucose metabolism under hypoxia in glioblastoma.

The impact of large language models on radiology: a guide for radiologists on the latest innovations in AI.

The advent of Deep Learning (DL) has significantly propelled the field of diagnostic radiology forward by enhancing image analysis and interpretation. The introduction of the Transformer architecture, followed by the development of Large Language Models (LLMs), has further revolutionized this domain. LLMs now possess the potential to automate and refine the radiology workflow, extending from report generation to assistance in diagnostics and patient care. The integration of multimodal technology with LLMs could potentially leapfrog these applications to unprecedented levels.However, LLMs come with unresolved challenges such as information hallucinations and biases, which can affect clinical reliability. Despite these issues, the legislative and guideline frameworks have yet to catch up with technological advancements. Radiologists must acquire a thorough understanding of these technologies to leverage LLMs' potential to the fullest while maintaining medical safety and ethics. This review aims to aid in that endeavor.

Machine learning-based gait analysis to predict clinical frailty scale in elderly patients with heart failure.

Aims: Although frailty assessment is recommended for guiding treatment strategies and outcome prediction in elderly patients with heart failure (HF), most frailty scales are subjective, and the scores vary among raters. We sought to develop a machine learning-based automatic rating method/system/model of the clinical frailty scale (CFS) for patients with HF. Methods and results: We prospectively examined 417 elderly (≥75 years) with symptomatic chronic HF patients from 7 centres between January 2019 and October 2023. The patients were divided into derivation (n = 194) and validation (n = 223) cohorts. We obtained body-tracking motion data using a deep learning-based pose estimation library, on a smartphone camera. Predicted CFS was calculated from 128 key features, including gait parameters, using the light gradient boosting machine (LightGBM) model. To evaluate the performance of this model, we calculated Cohen's weighted kappa (CWK) and intraclass correlation coefficient (ICC) between the predicted and actual CFSs. In the derivation and validation datasets, the LightGBM models showed excellent agreements between the actual and predicted CFSs [CWK 0.866, 95% confidence interval (CI) 0.807-0.911; ICC 0.866, 95% CI 0.827-0.898; CWK 0.812, 95% CI 0.752-0.868; ICC 0.813, 95% CI 0.761-0.854, respectively]. During a median follow-up period of 391 (inter-quartile range 273-617) days, the higher predicted CFS was independently associated with a higher risk of all-cause death (hazard ratio 1.60, 95% CI 1.02-2.50) after adjusting for significant prognostic covariates. Conclusion: Machine learning-based algorithms of automatically CFS rating are feasible, and the predicted CFS is associated with the risk of all-cause death in elderly patients with HF.

Intracerebral transplantation of MRI-trackable autologous bone marrow stromal cells for patients with subacute ischemic stroke.

BACKGROUND: Ischemic stroke is one of the leading causes of death and neurological disability worldwide, and stem cell therapy is highly expected to reverse the sequelae. This phase 1/2, first-in-human study evaluated the safety, feasibility, and monitoring of an intracerebral-transplanted magnetic resonance imaging (MRI)-trackable autologous bone marrow stromal cell (HUNS001-01) for patients with subacute ischemic stroke. METHODS: The study included adults with severe disability due to ischemic stroke. HUNS001-01 cultured with human platelet lysates and labeled with superparamagnetic iron oxide was stereotactically transplanted into the peri-infarct area 47-64 days after ischemic stroke onset (dose: 2 or 5 × 107 cells). Neurological and radiographic evaluations were performed throughout 1 year after cell transplantation. The trial was registered at UMIN Clinical Trial Registry (number UMIN000026130). FINDINGS: All seven patients who met the inclusion criteria successfully achieved cell expansion, underwent intracerebral transplantation, and completed 1 year of follow-up. No product-related adverse events were observed. The median National Institutes of Health Stroke Scale and modified Rankin scale scores before transplantation were 13 and 4, which showed improvements of 1-8 and 0-2, respectively. Cell tracking proved that the engrafted cells migrated toward the infarction border area 1-6 months after transplantation, and the quantitative susceptibility mapping revealed that cell signals at the migrated area constantly increased throughout the follow-up period up to 34% of that of the initial transplanted site. CONCLUSIONS: Intracerebral transplantation of HUNS001-01 was safe and well tolerated. Cell tracking shed light on the therapeutic mechanisms of intracerebral transplantation. FUNDING: This work was supported by the Japan Agency for Medical Research and Development (AMED; JP17bk0104045 and JP20bk0104011).

A Review of Hypoxia Imaging Using 18F-Fluoromisonidazole Positron Emission Tomography.

Tumor hypoxia is an essential factor related to malignancy, prognosis, and resistance to treatment. Positron emission tomography (PET) is a modality that visualizes the distribution of radiopharmaceuticals administered into the body. PET imaging with [18F]fluoromisonidazole ([18F]FMISO) identifies hypoxic tissues. Unlike [18F]fluorodeoxyglucose ([18F]FDG)-PET, fasting is not necessary for [18F]FMISO-PET, but the waiting time from injection to image acquisition needs to be relatively long (e.g., 2-4 h). [18F]FMISO-PET images can be displayed on an ordinary commercial viewer on a personal computer (PC). While visual assessment is fundamental, various quantitative indices such as tumor-to-muscle ratio have also been proposed. Several novel hypoxia tracers have been invented to compensate for the limitations of [18F]FMISO.

Nuclear medicine practice in Japan: a report of the ninth nationwide survey in 2022.

Subcommittee on Survey of Nuclear Medicine Practice in Japan has performed a nationwide survey of nuclear medicine practice every 5 years since 1982 to survey contemporary nuclear medicine practice and its changes over the years. The subcommittee sent questionnaires, including the number and category of examinations as well as the kind of the radiopharmaceuticals during the 30 days of June 2022 to all nuclear medicine institutes in Japan. The total numbers of them for the year 2022 were estimated depends on the 1-month data. A total of 1095 institutes responded to the survey, including 364 positron emission tomography (PET) centers. The recovery rate was 90.6%. The number of gamma cameras installed was 1299 in total, with 2.5% decrease in 5 years. Dual-head cameras and hybrid SPECT/CT scanners accounted for 83.8% and 35.5%, respectively. The number of single-photon tracer studies in 2022 was 1.11 million which means increase in 2.7% in 5 years. Bone scintigraphy was a leading examination (31.0%), followed by myocardial scintigraphy (27.1%) and cerebral perfusion study (23.8%) in order. The percentage of SPECT studies showed an increase from 63.5% in previous survey to 66.8% in this survey. PET centers have also increased from 389 to 412, as compared with the previous one. One hundred and twenty-two PET centers have installed one or two in-house cyclotrons. Increasing trends of the PET studies were observed from 1992 to 2017, the trend changed and PET studies showed 1.5% decrease in 5 years. 18F-FDG accounted for 98.6% (610,497 examinations). PET examinations using 11C-methionine, 13N-NH3 and 11C-PIB have decreased, with 1624, 2146 and 525 examinations, respectively in 2022. The total number of nuclear medicine examination was eventually increased by 1.0%. Therapies for pheochromocytoma or paraganglioma (PPGL) with 131I-MIBG and for neuroendocrine tumor with 177Lu-DOTA-TATE were newly started, however, a total number of targeted radionuclide therapy was decreased by 17.7% because 131I-radioiodine and 223Ra targeted therapies were decreased and supply of some radioisotopes was discontinued. 131I-radioiodine targeted therapy showed a decrease in 5 years (- 15.9%), including 4099 patients for thyroid cancer. The number of out-patient thyroid bed ablation therapy with 1110 MBq of 131I was also decreased to 1015 per year. The number of admission rooms specialized for radionuclide targeted therapy increased from 157 to 160. The number of 223Ra targeted therapies for castration-resistant metastatic prostate cancer (mCRPC) was 1041 patients. This survey was performed during COVID-19 pandemic, however, total number of nuclear medicine examinations was almost same as previous survey (+ 1.0%). Radionuclide therapies with 131I-MIBG and 177Lu-DOTA-TATE were newly started, and new radionuclide therapy will be available in future, therefore, the development of radionuclide therapy will be continued. We are convinced that this survey report is useful in understanding the current status of the nuclear medicine practice in Japan, and in devising the new strategy to strengthen a role of nuclear medicine.

Diagnosis of skull-base invasion by nasopharyngeal tumors on CT with a deep-learning approach.

PURPOSE: To develop a convolutional neural network (CNN) model to diagnose skull-base invasion by nasopharyngeal malignancies in CT images and evaluate the model's diagnostic performance. MATERIALS AND METHODS: We divided 100 malignant nasopharyngeal tumor lesions into a training (n = 70) and a test (n = 30) dataset. Two head/neck radiologists reviewed CT and MRI images and determined the positive/negative skull-base invasion status of each case (training dataset: 29 invasion-positive and 41 invasion-negative; test dataset: 13 invasion-positive and 17 invasion-negative). Preprocessing involved extracting continuous slices of the nasopharynx and clivus. The preprocessed training dataset was used for transfer learning with Residual Neural Networks 50 to create a diagnostic CNN model, which was then tested on the preprocessed test dataset to determine the invasion status and model performance. Original CT images from the test dataset were reviewed by a radiologist with extensive head/neck imaging experience (senior reader: SR) and another less-experienced radiologist (junior reader: JR). Gradient-weighted class activation maps (Grad-CAMs) were created to visualize the explainability of the invasion status classification. RESULTS: The CNN model's diagnostic accuracy was 0.973, significantly higher than those of the two radiologists (SR: 0.838; JR: 0.595). Receiver operating characteristic curve analysis gave an area under the curve of 0.953 for the CNN model (versus 0.832 and 0.617 for SR and JR; both p < 0.05). The Grad-CAMs suggested that the invasion-negative cases were present predominantly in bone marrow, while the invasion-positive cases exhibited osteosclerosis and nasopharyngeal masses. CONCLUSIONS: This CNN technique would be useful for CT-based diagnosis of skull-base invasion by nasopharyngeal malignancies.

Fairness of artificial intelligence in healthcare: review and recommendations.

In this review, we address the issue of fairness in the clinical integration of artificial intelligence (AI) in the medical field. As the clinical adoption of deep learning algorithms, a subfield of AI, progresses, concerns have arisen regarding the impact of AI biases and discrimination on patient health. This review aims to provide a comprehensive overview of concerns associated with AI fairness; discuss strategies to mitigate AI biases; and emphasize the need for cooperation among physicians, AI researchers, AI developers, policymakers, and patients to ensure equitable AI integration. First, we define and introduce the concept of fairness in AI applications in healthcare and radiology, emphasizing the benefits and challenges of incorporating AI into clinical practice. Next, we delve into concerns regarding fairness in healthcare, addressing the various causes of biases in AI and potential concerns such as misdiagnosis, unequal access to treatment, and ethical considerations. We then outline strategies for addressing fairness, such as the importance of diverse and representative data and algorithm audits. Additionally, we discuss ethical and legal considerations such as data privacy, responsibility, accountability, transparency, and explainability in AI. Finally, we present the Fairness of Artificial Intelligence Recommendations in healthcare (FAIR) statement to offer best practices. Through these efforts, we aim to provide a foundation for discussing the responsible and equitable implementation and deployment of AI in healthcare.

Comparative study of physiological FDG uptake in small structures between silicon photomultiplier-based PET and conventional PET.

OBJECTIVE: Silicon photomultiplier-based positron emission tomography/computed tomography (SiPM-PET/CT) has the superior spatial resolution to conventional PET/CT (cPET/CT). This head-to-head comparison study compared the images of physiological 18F-fluorodeoxyglucose (FDG) accumulation in small-volume structures between SiPM-PET/CT and cPET/CT in patients scanned with both modalities, and we investigated whether the thresholds that are reported to be useful for differentiating physiological accumulations from malignant lesions can also be applied to SiPM-PET/CT. METHODS: We enrolled 21 consecutive patients with head and neck malignancies who underwent whole-body FDG-PET/CT for initial staging or a follow-up evaluation (October 2020 to March 2022). After being injected with FDG, all patients underwent PET acquisition on both Vereos PET-CT and Gemini TF64 PET-CT systems (both Philips Healthcare) in random order. For each patient, the maximum standardized uptake value (SUVmax) was measured in the pituitary gland, esophagogastric junction (EGJ), adrenal glands, lumbar enlargement of the spinal cord, and epididymis. We measured the liver SUVmean and the blood pool SUVmean to calculate the target-to-liver ratio (TLR) and the target-to-blood ratio (TBR), respectively. Between-groups differences in each variable were examined by a paired t-test. We also investigated whether there were cases of target uptake greater than the reported threshold for distinguishing pathological from physiological accumulations. RESULTS: Data were available for 19 patients. Ten patients were in Group 1, i.e., the patients who underwent SiPM-PET first, and the remaining nine patients who underwent cPET first were in Group 2. In the SiPM-PET results, the SUVmax of all targets was significantly higher than that obtained by cPET in all patients, and this tendency was also observed when the patients were divided into Groups 1/2. The TLRs of all targets were significantly higher in SiPM-PET than in cPET in all patients, and SiPM-PET also showed significantly higher TBRs for all targets except the EGJ (p = 0.052). CONCLUSIONS: The physiological uptake in the small structures studied herein showed high accumulation on SiPM-PET. Our results also suggest that the thresholds reported for cPET to distinguish pathological accumulations likely lead to false-positive findings in SIPM-PET evaluations.

Revolutionizing radiation therapy: the role of AI in clinical practice.

This review provides an overview of the application of artificial intelligence (AI) in radiation therapy (RT) from a radiation oncologist's perspective. Over the years, advances in diagnostic imaging have significantly improved the efficiency and effectiveness of radiotherapy. The introduction of AI has further optimized the segmentation of tumors and organs at risk, thereby saving considerable time for radiation oncologists. AI has also been utilized in treatment planning and optimization, reducing the planning time from several days to minutes or even seconds. Knowledge-based treatment planning and deep learning techniques have been employed to produce treatment plans comparable to those generated by humans. Additionally, AI has potential applications in quality control and assurance of treatment plans, optimization of image-guided RT and monitoring of mobile tumors during treatment. Prognostic evaluation and prediction using AI have been increasingly explored, with radiomics being a prominent area of research. The future of AI in radiation oncology offers the potential to establish treatment standardization by minimizing inter-observer differences in segmentation and improving dose adequacy evaluation. RT standardization through AI may have global implications, providing world-standard treatment even in resource-limited settings. However, there are challenges in accumulating big data, including patient background information and correlating treatment plans with disease outcomes. Although challenges remain, ongoing research and the integration of AI technology hold promise for further advancements in radiation oncology.

Convolutional neural network-based program to predict lymph node metastasis of non-small cell lung cancer using 18F-FDG PET.

PURPOSE: To develop a convolutional neural network (CNN)-based program to analyze maximum intensity projection (MIP) images of 2-deoxy-2-[F-18]fluoro-D-glucose (FDG) positron emission tomography (PET) scans, aimed at predicting lymph node metastasis of non-small cell lung cancer (NSCLC), and to evaluate its effectiveness in providing diagnostic assistance to radiologists. METHODS: We obtained PET images of NSCLC from public datasets, including those of 435 patients with available N-stage information, which were divided into a training set (n = 304) and a test set (n = 131). We generated 36 maximum intensity projection (MIP) images for each patient. A residual network (ResNet-50)-based CNN was trained using the MIP images of the training set to predict lymph node metastasis. Lymph node metastasis in the test set was predicted by the trained CNN as well as by seven radiologists twice: first without and second with CNN assistance. Diagnostic performance metrics, including accuracy and prediction error (the difference between the truth and the predictions), were calculated, and reading times were recorded. RESULTS: In the test set, 67 (51%) patients exhibited lymph node metastases and the CNN yielded 0.748 predictive accuracy. With the assistance of the CNN, the prediction error was significantly reduced for six of the seven radiologists although the accuracy did not change significantly. The prediction time was significantly reduced for five of the seven radiologists with the median reduction ratio 38.0%. CONCLUSION: The CNN-based program could potentially assist radiologists in predicting lymph node metastasis by increasing diagnostic confidence and reducing reading time without affecting diagnostic accuracy, at least in the limited situations using MIP images.

Bone Echinococcosis Mimicking Malignancy on FDG PET.

ABSTRACT: MRI revealed a thoracic vertebrae lesion in a 40-year-old woman with back pain. She was referred to our institution; MRI demonstrated a mass from the second to the fifth thoracic vertebra and compression fractures. CT revealed a splenic mass, multiple pulmonary nodules, and low-density masses in the liver. 18 F-FDG PET/CT showed increased uptake (SUV max , 10.6) in the peripheral rim of the thoracic vertebra mass, with central parts showing lower uptake than the peripheral rim. The splenic mass exhibited increased accumulation (SUV max , 4.8). The thoracic spine lesion was fixed; a biopsy was performed. Alveolar echinococcosis was confirmed immunologically. Alveolar echinococcosis can present with bone lesions. It must be differentiated from malignancy.