Predicting subgroup treatment effects for a new study: Motivations, results and learnings from running a data challenge in a pharmaceutical corporation.

We present the motivation, experience, and learnings from a data challenge conducted at a large pharmaceutical corporation on the topic of subgroup identification. The data challenge aimed at exploring approaches to subgroup identification for future clinical trials. To mimic a realistic setting, participants had access to 4 Phase III clinical trials to derive a subgroup and predict its treatment effect on a future study not accessible to challenge participants. A total of 30 teams registered for the challenge with around 100 participants, primarily from Biostatistics organization. We outline the motivation for running the challenge, the challenge rules, and logistics. Finally, we present the results of the challenge, the participant feedback as well as the learnings. We also present our view on the implications of the results on exploratory analyses related to treatment effect heterogeneity.

Coherent, time-shifted patterns of microstructural plasticity during motor-skill learning.

Motor skill learning relies on neural plasticity in the motor and limbic systems. However, the spatial and temporal characteristics of these changes-and their microstructural underpinnings-remain unclear. Eighteen healthy males received 1 h of training in a computer-based motion game, 4 times a week, for 4 consecutive weeks, while 14 untrained participants underwent scanning only. Performance improvements were observed in all trained participants. Serial myelin- and iron-sensitive multiparametric mapping at 3T during this period of intensive motor skill acquisition revealed temporally and spatially distributed, performance-related microstructural changes in the grey and white matter across a corticospinal-cerebellar-hippocampal circuit. Analysis of the trajectory of these transient changes suggested time-shifted cascades of plasticity from the dominant sensorimotor system to the contralateral hippocampus. In the cranial corticospinal tracts, changes in myelin-sensitive metrics during training in the posterior limb of the internal capsule were of greater magnitude in those who trained their upper limbs vs. lower limb trainees. Motor skill learning is associated with waves of grey and white matter plasticity, across a broad sensorimotor network.

Simultaneous voxel-wise analysis of brain and spinal cord morphometry and microstructure within the SPM framework.

To validate a simultaneous analysis tool for the brain and cervical cord embedded in the statistical parametric mapping (SPM) framework, we compared trauma-induced macro- and microstructural changes in spinal cord injury (SCI) patients to controls. The findings were compared with results obtained from existing processing tools that assess the brain and spinal cord separately. A probabilistic brain-spinal cord template (BSC) was generated using a generative semi-supervised modelling approach. The template was incorporated into the pre-processing pipeline of voxel-based morphometry and voxel-based quantification analyses in SPM. This approach was validated on T1-weighted scans and multiparameter maps, by assessing trauma-induced changes in SCI patients relative to controls and comparing the findings with the outcome from existing analytical tools. Consistency of the MRI measures was assessed using intraclass correlation coefficients (ICC). The SPM approach using the BSC template revealed trauma-induced changes across the sensorimotor system in the cord and brain in SCI patients. These changes were confirmed with established approaches covering brain or cord, separately. The ICC in the brain was high within regions of interest, such as the sensorimotor cortices, corticospinal tracts and thalamus. The simultaneous voxel-wise analysis of brain and cervical spinal cord was performed in a unique SPM-based framework incorporating pre-processing and statistical analysis in the same environment. Validation based on a SCI cohort demonstrated that the new processing approach based on the brain and cord is comparable to available processing tools, while offering the advantage of performing the analysis simultaneously across the neuraxis.

Microstructural plasticity in nociceptive pathways after spinal cord injury.

OBJECTIVE: To track the interplay between (micro-) structural changes along the trajectories of nociceptive pathways and its relation to the presence and intensity of neuropathic pain (NP) after spinal cord injury (SCI). METHODS: A quantitative neuroimaging approach employing a multiparametric mapping protocol was used, providing indirect measures of myelination (via contrasts such as magnetisation transfer (MT) saturation, longitudinal relaxation (R1)) and iron content (via effective transverse relaxation rate (R2*)) was used to track microstructural changes within nociceptive pathways. In order to characterise concurrent changes along the entire neuroaxis, a combined brain and spinal cord template embedded in the statistical parametric mapping framework was used. Multivariate source-based morphometry was performed to identify naturally grouped patterns of structural variation between individuals with and without NP after SCI. RESULTS: In individuals with NP, lower R1 and MT values are evident in the primary motor cortex and dorsolateral prefrontal cortex, while increases in R2* are evident in the cervical cord, periaqueductal grey (PAG), thalamus and anterior cingulate cortex when compared with pain-free individuals. Lower R1 values in the PAG and greater R2* values in the cervical cord are associated with NP intensity. CONCLUSIONS: The degree of microstructural changes across ascending and descending nociceptive pathways is critically implicated in the maintenance of NP. Tracking maladaptive plasticity unravels the intimate relationships between neurodegenerative and compensatory processes in NP states and may facilitate patient monitoring during therapeutic trials related to pain and neuroregeneration.

Simultaneous assessment of regional distributions of atrophy across the neuraxis in MS patients.

BACKGROUND: The ability to assess brain and cord atrophy simultaneously would improve the efficiency of MRI to track disease evolution. OBJECTIVE: To test a promising tool to simultaneously map the regional distribution of atrophy in multiple sclerosis (MS) patients across the brain and cord. METHODS: Voxel-based morphometry combined with a statistical parametric mapping probabilistic brain-spinal cord (SPM-BSC) template was applied to standard T1-weighted magnetic resonance imaging (MRI) scans covering the brain and cervical cord from 37 MS patients and 20 healthy controls (HC). We also measured the cord area at C2-C3 with a semi-automatic segmentation method using (i) the same T1-weighted acquisitions used for the new voxel-based analysis and (ii) dedicated spinal cord phase sensitive inversion recovery (PSIR) acquisitions. Cervical cord findings derived from the three approaches were compared to each other and the goodness to fit to clinical scores was assessed by regression analyses. RESULTS: The SPM-BSC approach revealed a severity-dependent pattern of atrophy across the cervical cord and thalamus in MS patients when compared to HCs. The magnitude of cord atrophy was confirmed by the semi-automatic extraction approach at C2-C3 using both standard brain T1-weighted and advanced cord dedicated acquisitions. Associations between atrophy of cord and thalamus with disability and cognition were demonstrated. CONCLUSION: Atrophy in the brain and cervical cord of MS patients can be identified simultaneously and rapidly at the voxel-level. The SPM-BSC approach yields similar results as available standard processing tools with the added advantage of performing the analysis simultaneously and faster.

Model-based multi-parameter mapping.

Quantitative MR imaging is increasingly favoured for its richer information content and standardised measures. However, computing quantitative parameter maps, such as those encoding longitudinal relaxation rate (R1), apparent transverse relaxation rate (R2*) or magnetisation-transfer saturation (MTsat), involves inverting a highly non-linear function. Many methods for deriving parameter maps assume perfect measurements and do not consider how noise is propagated through the estimation procedure, resulting in needlessly noisy maps. Instead, we propose a probabilistic generative (forward) model of the entire dataset, which is formulated and inverted to jointly recover (log) parameter maps with a well-defined probabilistic interpretation (e.g., maximum likelihood or maximum a posteriori). The second order optimisation we propose for model fitting achieves rapid and stable convergence thanks to a novel approximate Hessian. We demonstrate the utility of our flexible framework in the context of recovering more accurate maps from data acquired using the popular multi-parameter mapping protocol. We also show how to incorporate a joint total variation prior to further decrease the noise in the maps, noting that the probabilistic formulation allows the uncertainty on the recovered parameter maps to be estimated. Our implementation uses a PyTorch backend and benefits from GPU acceleration. It is available at https://github.com/balbasty/nitorch.

Tracking the neurodegenerative gradient after spinal cord injury.

OBJECTIVE: To quantify neurodegenerative changes along the cervical spinal cord rostral to a spinal cord injury (SCI) by means of quantitative MRI (qMRI) and to determine its relationship with clinical impairment. METHODS: Thirty chronic SCI patients (15 tetraplegics and 15 paraplegics) and 23 healthy controls underwent a high-resolution T1-weighted and myelin-sensitive magnetization transfer (MT) MRI. We assessed macro- and microstructural changes along the cervical cord from levels C1 to C4, calculating cross-sectional spinal cord area, its anterior-posterior and left-right width and myelin content (i.e. MT). Regression analysis determined associations between qMRI parameters and clinical impairment. RESULTS: In SCI patients, cord area decreased by 2.67 mm2 (p = 0.004) and left-right width decreased by 0.35 mm (p = 0.002) per cervical cord level in the caudal direction when compared to the healthy controls. This gradient of neurodegeneration was greater in tetraplegic than paraplegics in the cross-sectional cervical cord area (by 3.28 mm2, p = 0.011), left-right width (by 0.36 mm, p = 0.03), and mean cord MT (by 0.13%, p = 0.04), but independant of lesion severity (p > 0.05). Higher lesion level was associated with greater magnitudes of neurodegeneration. Greater loss in myelin content in the dorsal columns and spinothalamic tract was associated with worse light touch (p = 0.016) and pin prick score (p = 0.024), respectively. CONCLUSIONS: A gradient of neurodegeneration is evident in the cervical cord remote from a SCI. Tract-specific associations with appropriate clinical outcomes highlight that remote neurodegenerative changes are clinically eloquent. Monitoring the neurodegenerative gradient could be used to track treatment effects of regenerative and neuroprotective agents, both in trials targeting cervical and thoracic SCI patients.

Slow conduction in mixed cultured strands of primary ventricular cells and stem cell-derived cardiomyocytes.

Modern concepts for the treatment of myocardial diseases focus on novel cell therapeutic strategies involving stem cell-derived cardiomyocytes (SCMs). However, functional integration of SCMs requires similar electrophysiological properties as primary cardiomyocytes (PCMs) and the ability to establish intercellular connections with host myocytes in order to contribute to the electrical and mechanical activity of the heart. The aim of this project was to investigate the properties of cardiac conduction in a co-culture approach using SCMs and PCMs in cultured cell strands. Murine embryonic SCMs were pooled with fetal ventricular cells and seeded in predefined proportions on microelectrode arrays to form patterned strands of mixed cells. Conduction velocity (CV) was measured during steady state pacing. SCM excitability was estimated from action potentials measured in single cells using the patch clamp technique. Experiments were complemented with computer simulations of conduction using a detailed model of cellular architecture in mixed cell strands. CV was significantly lower in strands composed purely of SCMs (5.5 ± 1.5 cm/s, n = 11) as compared to PCMs (34.9 ± 2.9 cm/s, n = 21) at similar refractoriness (100% SCMs: 122 ± 25 ms, n = 9; 100% PCMs: 139 ± 67 ms, n = 14). In mixed strands combining both cell types, CV was higher than in pure SCMs strands, but always lower than in 100% PCM strands. Computer simulations demonstrated that both intercellular coupling and electrical excitability limit CV. These data provide evidence that in cultures of murine ventricular cardiomyocytes, SCMs cannot restore CV to control levels resulting in slow conduction, which may lead to reentry circuits and arrhythmias.

Quantifying progression of multiple sclerosis via classification of depth videos.

This paper presents new learning-based techniques for measuring disease progression in Multiple Sclerosis (MS) patients. Our system aims to augment conventional neurological examinations by adding quantitative evidence of disease progression. An off-the-shelf depth camera is used to image the patient at the examination, during which he/she is asked to perform carefully selected movements. Our algorithms then automatically analyze the videos, assessing the quality of each movement and classifying them as healthy or non-healthy. Our contribution is three-fold: We i) introduce ensembles of randomized SVM classifiers and compare them with decision forests on the task of depth video classification; ii) demonstrate automatic selection of discriminative landmarks in the depth videos, showing their clinical relevance; iii) validate our classification algorithms quantitatively on a new dataset of 1041 videos of both MS patients and healthy volunteers. We achieve average Dice scores well in excess of the 80% mark, confirming the validity of our approach in practical applications. Our results suggest that this technique could be fruitful for depth-camera supported clinical assessments for a range of conditions.