Machine learning for automated and real-time two-dimensional to three-dimensional registration of the spine using a single radiograph.

OBJECTIVE: The goal of this work was to methodically evaluate, optimize, and validate a self-supervised machine learning algorithm capable of real-time automatic registration and fluoroscopic localization of the spine using a single radiograph or fluoroscopic frame. METHODS: The authors propose a two-dimensional to three-dimensional (2D-3D) registration algorithm that maximizes an image similarity metric between radiographic images to identify the position of a C-arm relative to a 3D volume. This work utilizes digitally reconstructed radiographs (DRRs), which are synthetic radiographic images generated by simulating the x-ray projections as they would pass through a CT volume. To evaluate the algorithm, the authors used cone-beam CT data for 127 patients obtained from an open-source de-identified registry of cervical, thoracic, and lumbar scans. They systematically evaluated and tuned the algorithm, then quantified the convergence rate of the model by simulating C-arm registrations with 80 randomly simulated DRRs for each CT volume. The endpoints of this study were time to convergence, accuracy of convergence for each of the C-arm's degrees of freedom, and overall registration accuracy based on a voxel-by-voxel measurement. RESULTS: A total of 10,160 unique radiographic images were simulated from 127 CT scans. The algorithm successfully converged to the correct solution 82% of the time with an average of 1.96 seconds of computation. The radiographic images for which the algorithm converged to the solution demonstrated 99.9% registration accuracy despite utilizing only single-precision computation for speed. The algorithm was found to be optimized for convergence when the search space was limited to a ± 45° offset in the right anterior oblique/left anterior oblique, cranial/caudal, and receiver rotation angles with the radiographic isocenter contained within 8000 cm3 of the volumetric center of the CT volume. CONCLUSIONS: The investigated machine learning algorithm has the potential to aid surgeons in level localization, surgical planning, and intraoperative navigation through a completely automated 2D-3D registration process. Future work will focus on algorithmic optimizations to improve the convergence rate and speed profile.

Variability and heritability of mouse brain structure: Microscopic MRI atlases and connectomes for diverse strains.

Genome-wide association studies have demonstrated significant links between human brain structure and common DNA variants. Similar studies with rodents have been challenging because of smaller brain volumes. Using high field MRI (9.4 T) and compressed sensing, we have achieved microscopic resolution and sufficiently high throughput for rodent population studies. We generated whole brain structural MRI and diffusion connectomes for four diverse isogenic lines of mice (C57BL/6J, DBA/2J, CAST/EiJ, and BTBR) at spatial resolution 20,000 times higher than human connectomes. We measured narrow sense heritability (h2) I.e. the fraction of variance explained by strains in a simple ANOVA model for volumes and scalar diffusion metrics, and estimates of residual technical error for 166 regions in each hemisphere and connectivity between the regions. Volumes of discrete brain regions had the highest mean heritability (0.71 ± 0.23 SD, n = 332), followed by fractional anisotropy (0.54 ± 0.26), radial diffusivity (0.34 ± 0.022), and axial diffusivity (0.28 ± 0.19). Connection profiles were statistically different in 280 of 322 nodes across all four strains. Nearly 150 of the connection profiles were statistically different between the C57BL/6J, DBA/2J, and CAST/EiJ lines. Microscopic whole brain MRI/DTI has allowed us to identify significant heritable phenotypes in brain volume, scalar DTI metrics, and quantitative connectomes.

Pre-HAART CD4+ T-lymphocytes as biomarkers of post-HAART immune recovery in HIV-infected children with or without TB co-infection.

BACKGROUND: Infection with the Human Immunodeficiency Virus (HIV) dramatically increases the risk of developing active tuberculosis (TB). Several studies have indicated that co-infection with TB increases the risk of HIV progression and death. Sub-Saharan Africa bears the brunt of these dual epidemics, with about 2.4 million HIV-infected people living with TB. The main objective of our study was to assess whether the pre-HAART CD4+ T-lymphocyte counts and percentages could serve as biomarkers for post-HAART treatment immune-recovery in HIV-positive children with and without TB co-infection. METHODS: The data analyzed in this retrospective study were collected from a cohort of 305 HIV-infected children being treated with HAART. A Lehmann family of ROC curves were used to assess the diagnostic performance of pre- HAART treatment CD4+ T-lymphocyte count and percentage as biomarkers for post-HAART immune recovery. The Kaplan-Meier estimator was used to compare differences in post-HAART recovery times between patients with and without TB co-infection. RESULTS: We found that the diagnostic performance of both pre-HARRT treatment CD4+ T-lymphocyte count and percentage was comparable and achieved accuracies as high as 74%. Furthermore, the predictive capability of pre-HAART CD4+ T-lymphocyte count and percentage were slightly better in TB-negative patients. Our analyses also indicate that TB-negative patients have a shorter recovery time compared to the TB-positive patients. CONCLUSIONS: Pre-HAART CD4+ T-lymphocyte count and percentage are stronger predictors of immune recovery in TB-negative pediatric patients, suggesting that TB co-infection complicates the treatment of HIV in this cohort. These findings suggest that the detection and treatment of TB is essential for the effectiveness of HAART in HIV-infected pediatric patients.

A comparison of machine learning techniques for classification of HIV patients with antiretroviral therapy-induced mitochondrial toxicity from those without mitochondrial toxicity.

BACKGROUND: Antiretroviral therapy (ART) has significantly reduced HIV-related morbidity and mortality. However, therapeutic benefit of ART is often limited by delayed drug-associated toxicity. Nucleoside reverse transcriptase inhibitors (NRTIs) are the backbone of ART regimens. NRTIs compete with endogenous deoxyribonucleotide triphosphates (dNTPs) in incorporation into elongating DNA chain resulting in their cytotoxic or antiviral effect. Thus, the efficacy of NRTIs could be affected by direct competition with endogenous dNTPs and/or feedback inhibition of their metabolic enzymes. In this paper, we assessed whether the levels of ribonucleotides (RN) and dNTP pool sizes can be used as biomarkers in distinguishing between HIV-infected patients with ART-induced mitochondrial toxicity and HIV-infected patients without toxicity. METHODS: We used data collected through a case-control study from 50 subjects. Cases were defined as HIV-infected individuals with clinical and/or laboratory evidence of mitochondrial toxicity. Each case was age, gender, and race matched with an HIV-positive without evidence of toxicity. We used a range of machine learning procedures to distinguish between patients with and without toxicity. Using resampling methods like Monte Carlo k-fold cross validation, we compared the accuracy of several machine learning algorithms applied to our data. We used the algorithm with highest classification accuracy rate in evaluating the diagnostic performance of 12 RN and 14 dNTP pool sizes as biomarkers of mitochondrial toxicity. RESULTS: We used eight classification algorithms to assess the diagnostic performance of RN and dNTP pool sizes distinguishing HIV patients with and without NRTI-associated mitochondrial toxicity. The algorithms resulted in cross-validated classification rates of 0.65-0.76 for dNTP and 0.72-0.83 for RN, following reduction of the dimensionality of the input data. The reduction of input variables improved the classification performance of the algorithms, with the most pronounced improvement for RN. Complex tree-based methods worked the best for both the deoxyribose dataset (Random Forest) and the ribose dataset (Classification Tree and AdaBoost), but it is worth noting that simple methods such as Linear Discriminant Analysis and Logistic Regression were very competitive in terms of classification performance. CONCLUSIONS: Our finding of changes in RN and dNTP pools in participants with mitochondrial toxicity validates the importance of dNTP pools in mitochondrial function. Hence, levels of RN and dNTP pools can be used as biomarkers of ART-induced mitochondrial toxicity.