(113 days)
Bullsai is composed of a set of modules intended for analysis and processing of medical images and other healthcare data. It includes functions for image manipulation, basic measurements, and planning.
Bullsai is indicated for use in image processing, registration, atlas-assisted visualization and segmentation, and target export creation and selection of structural MRI images, where an output can be generated for use by a system capable of reading DICOM image sets.
Bullsai is indicated for use in the processing of diffusion-weighted MRI sequences into 3D maps that represent white matter tracts based on diffusion reconstruction methods and for the use of said maps to select and create exports.
Typical users of Bullsai are medical professionals, including but not limited to surgeons, clinicians. and radiologists.
Bullsai is a software-only, cloud-deployed, image processing package which can be used to perform DICOM image processing and analysis.
Bullsai can receive ("import") DICOM images from picture archiving and communication systems (PACS), including Diffusion Weighted Imaging (DWI) and structural brain imaging.
Bullsai removes protected health information (PHI) and links the dataset to an encryption key, which is then used to relink the data back to the patient when the data is exported to a medical imaging platform such as a hospital PACS or other DICOM device.
The software provides a workflow for a clinician to:
- Select an image for planning and visualization,
- Validate image quality,
- Export black and white and color DICOMs for use in systems that can view DICOM images.
Bullsai uses advanced MRI processing to deliver patient-specific anatomy and tractography results to support physicians in neurosurgical planning. Bullasi preprocessing steps include denoising, debiasing, skull stripping, susceptibility distortion and head motion correction to ensure input data can support generation of tractography and segmentation results. Bullsai uses generalized q-sampling imaging (GQI) to estimate the voxel-level white matter microstructure: GQI is a model-free diffusion reconstruction method that uses the Fourier transform relationship between the diffuse MR signal and the potential diffusion displacement for resolving fiber orientations and the anisotropy of water. Patient-specific anatomical segmentation and GQI estimated fiber orientations are used in Bullsai as part of an iterative heuristic tractography algorithm that emulates the manual work of neuroanatomists who iteratively seed tracts and remove aberrant fibers. Results are shared as DICOM outputs which can readily be viewed and edited in standard neurosurgical planning software packages.
The provided text contains details about the Bullsai device, its indications for use, and a comparison to a predicate device (Quicktome Software Suite). However, it does not explicitly detail the acceptance criteria for a study or the results of a specific study proving the device meets those criteria, nor does it provide information regarding sample sizes for test/training sets, data provenance, number/qualification of experts, adjudication methods, MRMC studies, or standalone performance metrics.
The "Performance Testing Summary" section indicates that "Software verification and validation testing were conducted. Documentation and relevant standards were provided." but does not elaborate on the specifics of these tests or their results against defined acceptance criteria.
The "Validation of differences compared to predicate devices" section mentions three validation activities:
- "Clinical accuracy of the Bullsai device outputs were evaluated and validated by clinical experts for the clinical intended uses of presurgical planning."
- "Bullsai device output's were validated compatibility with major neuronavigation software systems."
- "Bullsai device output's were validated for repeatability and reproducibility across major scanner manufacturer brands."
However, no further details are provided about these validations.
Therefore,Based on the provided text, the specific information requested in the prompt's numbered points cannot be fully extracted as it is not explicitly stated.
Here's a summary of what can be inferred or what is missing:
1. A table of acceptance criteria and the reported device performance:
* Acceptance Criteria: Not explicitly stated in the document.
* Reported Device Performance: Not explicitly stated in the document in measurable terms against acceptance criteria. The document mentions "Clinical accuracy of the Bullsai device outputs were evaluated and validated by clinical experts" and "Bullsai device output's were validated compatibility" and "Bullsai device output's were validated for repeatability and reproducibility," but no specific performance metrics or thresholds are provided.
2. Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective):
* Sample Size (Test Set): Not stated.
* Data Provenance: Not stated.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience):
* Number of Experts: Not stated, only "clinical experts" are mentioned.
* Qualifications of Experts: Not stated.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set:
* Adjudication Method: Not stated.
5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:
* MRMC Study: Not mentioned as being performed. The device description does not imply a human-in-the-loop diagnostic aid but rather image processing and output generation for other systems. The predicate device's differences section states, "Bullsai does not have a viewer and therefore a usability study was not conducted," which suggests a traditional MRMC study involving human reading performance might not be directly applicable or was not undertaken.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
* Standalone Performance: The document implies standalone validation for "clinical accuracy," "compatibility," and "repeatability and reproducibility." However, specific quantitative results from such a standalone study are not provided.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
* Type of Ground Truth: The document states "Clinical accuracy of the Bullsai device outputs were evaluated and validated by clinical experts." This suggests expert consensus or expert review served as the ground truth. No mention of pathology or outcomes data is made.
8. The sample size for the training set:
* Sample Size (Training Set): Not stated.
9. How the ground truth for the training set was established:
* Ground Truth (Training Set): Not stated. The document focuses on validation activities, not training methodologies or ground truth establishment for training data.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).