Camp, Jon J.

Abstract Background/Purpose Diagnosing wrist ligament injuries is challenging; early detection and treatment are important to prevent osteoarthritis progression. Interosseous proximity maps, a proxy measure for joint space, can be generated from volumetric imaging data and may provide important information about wrist health. Artificial intelligence (AI) could enhance accuracy of noninvasive diagnosis based on imaging-derived metrics. This work demonstrates feasibility of AI training using synthetic proximity map data generated from finite element models (FEMs). Methods Personalized wrist FEMs for two asymptomatic participants were created from four-dimensional computed tomography-derived anatomic and kinematic data. Monte Carlo sampling varied 22 ligament material properties and simulated 7,500 unique injury scenarios generating 9,000,000 labeled red, green, and blue (RGB) images of interosseous proximity vector fields from FEM-derived motions. Images were associated with 17 descriptive metrics, including gross wrist angles and bone surface pairs, and used to develop mixed-input convolutional neural networks (CNNs). Model performance was evaluated for identifying specific ligament injuries. Results Average area under receiver operating characteristic curve (AUROC) for CNNs was 0.757 across all injury types and kinematics. In a subset with clinically-relevant functional angles, the average AUROC was 0.824. Best-performing individual ligament AUROCs ranged from 0.807 to 0.999. Sensitivities and specificities exceeded 0.99 for some ligament injury simulations under specific wrist angles and bone surface pairs. Conclusion This study demonstrates the feasibility of using synthetic data from FEMs to train AI models for classifying wrist ligament injuries. Proximity-based RGB images may be a relevant biomarker of ligamentous injury.

Abstract Senescent cells accumulate with age and contribute to tissue dysfunction, yet their identification in tissues is challenging due to low abundance, heterogeneous phenotypes, and the lack of specific markers. Senescence-associated features span multiple subcellular compartments, including nuclear DNA damage foci, cytosolic protein changes, and perinuclear alterations, each requiring tailored detection strategies. To overcome these challenges, we developed SenoQuant ( https://github.com/HaamsRee/senoquant ), a versatile software designed for comprehensive, accurate, and unbiased spatial quantification and prediction of senescence markers across diverse tissue contexts. Utilizing AI models, SenoQuant enables precise nuclear and cytoplasmic segmentation and detection of senescence markers across low- and high-plex imaging modalities, applicable to cultured cells and tissue sections from mice and humans. The platform also supports custom AI models; for example, we built SenCeption, a proof-of-concept predictor of single-cell p21 status from DAPI-stained nuclei in human skin. Available as a free napari plugin, SenoQuant is widely accessible to researchers. By providing a unified approach to senescence analysis and prediction, SenoQuant opens new opportunities for exploring the complex biology of senescence and its impacts on aging and disease.

Objective Preeclampsia, a hypertensive pregnancy disorder, increases the risk for cardiovascular disease including coronary artery calcification (CAC) as women age. Controversy exists regarding the relationship between bone mineral density (BMD) and CAC in menopausal women. This study was designed to determine and compare the relationship between BMD and CAC in postmenopausal women with defined histories of either normotensive or preeclamptic pregnancies. Methods Women who gave birth in Olmsted County, MN between 1976 and 1982 were identified from their medical record as having a preeclamptic pregnancy (PE, n=40) and were age/parity‐matched with women having normotensive pregnancies (NP, n=40). The study was approved by the IRB at Mayo Clinic and all participants gave written informed consent. Quantitative Computed Tomography was used to assess vertebral (T7‐9) BMD and CAC. Statistical analyses included linear regression with general estimating equation (GEE) to correct for within‐subject correlations. Results Women averaged 35 years since the index pregnancy and were 60 years of age at the time of the study. CAC was significantly greater in women of the PE group (median (25 th , 75 th percentiles: 0 (0, 0.3) NP and 0 (0, 25) PE, P=0.026). There were no significant differences in cortical, trabecular or central BMD between PE and NP groups. However, there was a positive correlation between cortical BMD and CAC in the PE group (ρ=0.336, P=0.042) and a negative correlation between central BMD and CAC in the NP group (ρ=−0.392, P=0.017). Conclusions Although CAC was greater in women with a history PE, BMD did not differ from women with a history of NP. Associations between cortical and central BMD with CAC differed by pregnancy history suggesting that there may be underlying differences in calcium metabolism associated with PE. Support or Funding Information This work was funded by grants from the NIH AG44170, AG034676, UL1 TR000135 and the Mayo Foundation.

Flow cytometric (FC) immunophenotyping is critical but time-consuming in diagnosing minimal residual disease (MRD). We evaluated whether human-in-the-loop artificial intelligence (AI) could improve the efficiency of clinical laboratories in detecting MRD in chronic lymphocytic leukemia (CLL). We developed deep neural networks (DNN) that were trained on a 10-color CLL MRD panel from treated CLL patients, including DNN trained on the full cohort of 202 patients (F-DNN) and DNN trained on 138 patients with low-event cases (MRD < 1000 events) (L-DNN). A hybrid DNN approach was utilized, with F-DNN and L-DNN applied sequentially to cases. “Ground truth” classification of CLL MRD was confirmed by expert analysis. The hybrid DNN approach demonstrated an overall accuracy of 97.1% (95% CI: 84.7–99.9%) in an independent cohort of 34 unknown samples. When CLL cells were reported as a percentage of total white blood cells, there was excellent correlation between the DNN and expert analysis [r > 0.999; Passing–Bablok slope = 0.997 (95% CI: 0.988–0.999) and intercept = 0.001 (95% CI: 0.000–0.001)]. Gating time was dramatically reduced to 12 s/case by DNN from 15 min/case by the manual process. The proposed DNN demonstrated high accuracy in CLL MRD detection and significantly improved workflow efficiency. Additional clinical validation is needed before it can be fully integrated into the existing clinical laboratory practice.

Kidney stones are a common urologic condition with a high amount of recurrence. Recurrence depends on a multitude of factors the incidence of precursors to kidney stones, plugs, and plaques. One method of characterising the stone precursors is endoscopic assessment, though it is manual and time‐consuming. Deep learning has become a popular technique for semantic segmentation because of the high accuracy that has been demonstrated. The present Letter examined the efficacy of deep learning to segment the renal papilla, plaque, and plugs. A U‐Net model with ResNet‐34 encoder was tested; the Letter examined dropout (to avoid overtraining) and two different loss functions (to address the class imbalance problem. The models were then trained in 1666 images and tested on 185 images. The Jaccard‐cross‐entropy loss function was more effective than the focal loss function. The model with the dropout rate 0.4 was found to be more effective due to its generalisability. The model was largely successful at delineating the papilla. The model was able to correctly detect the plaques and plugs; however, small plaques were challenging. Deep learning was found to be applicable for segmentation of an endoscopic image for the papilla, plaque, and plug, with room for improvement.

Background Since 1972, patients with large nasal perforations, who were symptomatic, and who were not candidates for surgery, had the option of custom prosthetic closure at Mayo Clinic. Although septal prostheses have helped many patients, 27% of pre-1982 patients chose not to keep the prosthesis in place. Two-dimensional computed tomography (CT) sizing resulted in more of the patients choosing to retain the prosthesis. The introduction of three-dimensional (3-D) printing to the sizing process offered the potential of further improved retention by refinement in prosthesis fit. Objective To describe the fabrication of nasal septal prostheses by using 3-D printing for sizing and to compare the retention rate of 3-D-sized prostheses with those that used previous sizing methods. Methods Twenty-one consecutive patients who had placement of septal prostheses sized by using 3-D printed templates were studied. CT image data were used to print 3-D templates of the exact shape of the patient's septal perforation, and medical-grade silastic prostheses were fabricated to fit. In four cases, the 3-D printed template allowed preoperative surgical simulation. Metrics collected included prosthesis retention; symptoms, including intranasal crusting and epistaxis; and previous prosthetic closure failures. Results Twenty of the twenty-one patients had improvement in symptoms. The mean diameter of the perforations was 2.4 cm; the mean closure time by the end of the study period was 2.2 years. All but two patients chose to keep their prosthesis in place, for a retention rate of 90%. Seven patients with successful closure had failed previously with prior prosthesis sized without the current 3-D printing methodology. This 90% retention rate exceeded the previous rates before the introduction of 3-D sizing. Conclusion Sizing done by 3-D printing for prosthetic closure of nasal septal perforations resulted in a higher retention rate in helping patients with these most-challenging nasal septal perforations.

The spine is the most common site for secondary bone metastases, and clinical management for fractures is based on size and geometry of the defect. About 75% of the bone needs to be damaged before lesions are detectable, so clinical tools should measure changes in both geometry and material properties. We have developed an automated, user-friendly, Spine Cancer Assessment (SCA) image-based analysis method that builds on a platform designed for clinical practice providing failure characteristics of vertebrae. The objectives of this study were to (1) validate SCA predictions with experimental failure load outcomes; (2) evaluate the planning capabilities for prophylactic vertebroplasty procedures; and (3) investigate the effect of computed tomography (CT) protocols on predicted failure loads. Twenty-one vertebrae were randomly divided into two groups: (1) simulated defect without treatment (negative control) [ n = 9] and (2) with treatment [ n = 12]. Defects were created and a polymeric biomaterial was injected into the vertebrae in the treated-defect group. Spines were scanned, reconstructed with two algorithms, and analyzed for fracture loads. To virtually plan for prophylactic intervention, vertebrae with empty lesions were simulated to be augmented with either poly(methyl methacrylate) (PMMA) or a novel bone replacement copolymer, poly(propylene fumarate-co-caprolactone) [P(PF-co-CL)]. Axial rigidities were calculated from the CT images. Failure loads, determined from the cross section with the lowest axial rigidity, were compared with experimental values. Predicted loads correlated well with experimental outcomes ( R 2 = 0.73, p < 0.0001). Predictions from negative control specimens highly correlated with measured values ( R 2 = 0.90, p < 0.0001). Although a similar correlation was obtained using both algorithms, the smooth reconstruction (B30) tended to underestimate predicted failure loads by ∼50% compared with the ∼10% underestimate of the sharp reconstruction (B70). Percent increase in failure loads after virtual vertebroplasty showed a higher increase in samples with PMMA compared with those with copolymer. The SCA method developed in this study calculated failure loads from quantitative computed tomography scans in vertebrae with simulated metastatic lytic defects, with or without treatment, facilitating clinical applicability and providing more reliable guidelines for physicians to select appropriate treatment options. Furthermore, the analysis could accommodate augmentation planning procedures that aimed to determine the optimum material that would increase vertebral body failure load.

Spine intersegmental motion parameters and the resultant regional patterns may be useful for biomechanical classification of low back pain (LBP) as well as assessing the appropriate intervention strategy. Because of its availability and reasonable cost, two-dimensional (2D) fluoroscopy has great potential as a diagnostic and evaluative tool. However, the technique of quantifying intervertebral motion in the lumbar spine must be validated, and the sensitivity assessed. The purpose of this investigation was to (1) compare synchronous fluoroscopic and optoelectronic measures of intervertebral rotations during dynamic flexion–extension movements in vitro and (2) assess the effect of C-arm rotation to simulate off-axis patient alignment on intervertebral kinematics measures. Six cadaveric lumbar–sacrum specimens were dissected, and active marker optoelectronic sensors were rigidly attached to the bodies of L2–S1. Fluoroscopic sequences and optoelectronic kinematic data (0.15-mm linear, 0.17–0.20 deg rotational, accuracy) were obtained simultaneously. After images were obtained in a true sagittal plane, the image receptor was rotated in 5 deg increments (posterior oblique angulations) from 5 deg to 15 deg. Quantitative motion analysis (qma) software was used to determine the intersegmental rotations from the fluoroscopic images. The mean absolute rotation differences between optoelectronic values and dynamic fluoroscopic values were less than 0.5 deg for all the motion segments at each off-axis fluoroscopic rotation and were not significantly different (P > 0.05) for any of the off-axis rotations of the fluoroscope. Small misalignments of the lumbar spine relative to the fluoroscope did not introduce measurement variation in relative segmental rotations greater than that observed when the spine and fluoroscope were perpendicular to each other, suggesting that fluoroscopic measures of relative segmental rotation during flexion–extension are likely robust, even when patient alignment is not perfect.