Evaluating artificial intelligence software for delineating hemorrhage extent on CT brain imaging in stroke: AI delineation of ICH on CT

What will your doctor do with this information to get you 100% recovered? If nothing here gets you to that goal then it was useless research. The whole fucking point of stroke research is to get survivors 100% recovered! Since you don't know that, you're fired!

Evaluating artificial intelligence software for delineating hemorrhage extent on CT brain imaging in stroke: AI delineation of ICH on CT

• Adam Vacek
• Grant Mair, MD
• Philip White, MD
• 
  
• Malcolm Macleod, PhD
• Nikola Sprigg, DM
• Joanna M Wardlaw, MD
• Show all authors

Open AccessPublished:November 25, 2023DOI:https://doi.org/10.1016/j.jstrokecerebrovasdis.2023.107512

Abstract

Background

: The extent and distribution of intracranial hemorrhage (ICH) directly affects clinical management. Artificial intelligence (AI) software can detect and may delineate ICH extent on brain CT. We evaluated e-ASPECTS software (Brainomix Ltd.) performance for ICH delineation.

Methods

: We qualitatively assessed software delineation of ICH on CT using patients from six stroke trials. We assessed hemorrhage delineation in five compartments: lobar, deep, posterior fossa, intraventricular, extra-axial. We categorized delineation as excellent, good, moderate, or poor. We assessed quality of software delineation with number of affected compartments in univariate analysis (Kruskall-Wallis test) and ICH location using logistic regression (dependent variable: dichotomous delineation categories ‘excellent-good’ versus ‘moderate-poor’), and report odds ratios (OR) and 95 % confidence intervals (95 %CI).

Results

: From 651 patients with ICH (median age 75 years, 53 % male), we included 628 with assessable CTs. Software delineation of ICH extent was ‘excellent’ in 189/628 (30 %), ‘good’ in 255/628 (41 %), ‘moderate’ in 127/628 (20 %), and ‘poor’ in 57/628 cases (9 %). The quality of software delineation of ICH was better when fewer compartments were affected (Z = 3.61-6.27; p = 0.0063). Software delineation of ICH extent was more likely to be ‘excellent-good’ quality when lobar alone (OR = 1.56, 95 %CI = 0.97-2.53) but ‘moderate-poor’ with any intraventricular (OR = 0.56, 95 %CI = 0.39-0.81, p = 0.002) or any extra-axial (OR = 0.41, 95 %CI = 0.27-0.62, p<0.001) extension.

Conclusions

: Delineation of ICH extent on stroke CT scans by AI software was excellent or good in 71 % of cases but was more likely to over- or under-estimate extent when ICH was either more extensive, intraventricular, or extra-axial.

Biomimetic Nanoparticles as a Theranostic Tool for Traumatic Brain Injury

This would seem to be much much better at identifying specific areas of the brain that are damaged. Much better than CT or MRI scans which indirectly identify brain damage. Now we just need our stroke leadership to improve this vastly so results can come back in minutes rather than 24 hours. BUT NOTHING WILL OCCUR, WE HAVE NO STROKE LEADERSHIP, we have no one with any brains at all in stroke. Your children and grandchildren will be screwed when they have strokes unless we get survivors in charge.

Biomimetic Nanoparticles as a Theranostic Tool for Traumatic Brain Injury

 Assaf Zinger,* Sirena Soriano, Gherardo Baudo, Enrica De Rosa, Francesca Taraballi,* and Sonia Villapol* 

Traumatic brain injury (TBI) triggers both central and peripheral inflammatory responses. Existing pharmacological drugs are unable to effectively and quickly target the brain inflamed regions, setting up a major roadblock towards effective brain trauma treatments. Nanoparticles (NPs) have been used in multiple diseases as drug delivery tools with remarkable success due to their rapid diffusion and specificity in the target organ. Here, leukocyte-based biomimetic NPs are fabricated as a theranostic tool to directly access inflamed regions in a TBI mouse model. This NP systemic delivery is visualized using advanced in vivo imaging techniques, including intravital microscopy and in vivo imaging system. The results demonstrate selective targeting of NPs to the injured brain and increased NPs accumulation among the peripheral organs 24 h after TBI. Interestingly, increased microglial proliferation, decreased macrophage infiltration, and reduced brain lesion following the NPs treatments compared to sham vehicle-treated mice are also found. In summary, the results suggest that NPs represent a promising future theranostic tool for TBI treatment.

Icometrix announces ischaemic stroke solution

I wondered if it does white matter damage from the stroke. It does find white matter hyperintensities from this research.  

How does icobrain enhance your radiological reporting for MS?

 

Because if white matter damage is found then that requires different protocols to solve. And your doctor has none.  

The latest here:

Icometrix announces ischaemic stroke solution

In a press release, icometrix announced the addition of icobrain cva, a stroke solution, to the icobrain portfolio. According to the company, this announcement follows clearance from the US Food and Drug Administration (FDA) and CE-marking of its image processing software for the analysis and communication of the tissue perfusion state on computer tomography (CT) perfusion scans in patients with ischaemic stroke. 

According to icometrix, the icobrain cva is a fully-automated software solution for the quantitative assessment of tissue perfusion on CT. The company claims that icobrain cva reports the volume of the core and perfusion lesion by quantifying reduced cerebral blood flow, volume, and transit time. Additionally, icometrix states the report includes information on the correctness of the selected arterial input function and the quality of the output.  

The press release details that icobrain cva provides physicians with fast, fully automated, and state-of-the-art insights to support treatment decisions in acute ischaemic stroke. It is further stated that the automated assessment of tissue parameters in an acute clinical setting by icobrain cva will allow more patients to get the right treatment and can improve patient outcomes and care while increasing efficiency. 

“With the launch of icobrain cva we address a persisting need in the treatment of acute ischaemic stroke. By democratising advanced CT perfusion analysis for healthcare systems worldwide, we take the next step in our mission to become a holistic brain solution provider,” says Wim Van Hecke, CEO at icometrix, Antwerp, Belgium. 

“The main challenge of current stroke solutions is correctly identifying the entry point of the injected contrast in the brain. icobrain cva introduces new, patented, deep learning technology into this identification process to achieve a more robust assessment of the infarcted area,” states Dirk Smeets, CTO at icometrix, Leuven, Belgium.

 

Length of stay benchmarks for inpatient rehabilitation after stroke

From Canada but my reading of it does not inspire confidence. I didn't see any mention of using the MRI/CT scans to identify a damage diagnosis and no mention of using direct patient outcomes.
Unless we as survivors get involved in these the end result will not be useful for our recovery(I refuse to use the term compensation). 

Length of stay benchmarks for inpatient rehabilitation after stroke

Purpose: In Canada, no standardized benchmarks for length of stay (LOS) have been established for post-stroke inpatient rehabilitation. This paper describes the development of a severity specific median length of stay benchmarking strategy, assessment of its impact after one year of implementation in a Canadian rehabilitation hospital, and establishment of updated benchmarks that may be useful for comparison with other facilities across Canada. Method: Patient data were retrospectively assessed for all patients admitted to a single post-acute stroke rehabilitation unit in Ontario, Canada between April 2005 and March 2008. Rehabilitation Patient Groups (RPGs) were used to establish stratified median length of stay benchmarks for each group that were incorporated into team rounds beginning in October 2009. Benchmark impact was assessed using mean LOS, FIM^® gain, and discharge destination for each RPG group, collected prospectively for one year, compared against similar information from the previous calendar year. Benchmarks were then adjusted accordingly for future use. Results: Between October 2009 and September 2010, a significant reduction in average LOS was noted compared to the previous year (35.3 vs. 41.2 days; p < 0.05). Reductions in LOS were noted in each RPG group including statistically significant reductions in 4 of the 7 groups. As intended, reductions in LOS were achieved with no significant reduction in mean FIM^® gain or proportion of patients discharged home compared to the previous year. Adjusted benchmarks for LOS ranged from 13 to 48 days depending on the RPG group. Conclusions: After a single year of implementation, severity specific benchmarks helped the rehabilitation team reduce LOS while maintaining the same levels of functional gain and achieving the same rate of discharge to the community.

Implications for Rehabilitation

• Efficient post-stroke rehabilitation can help to improve patient outcomes and reduce the financial burden placed on the healthcare system.
• Yet, unnecessarily long lengths of stay in rehabilitation are not in the best interest of the patient and act to increase the cost of care.
• This study illustrates how a length of stay benchmarking system can help to promote efficiency in post-stroke rehabilitation and reduce the cost of care without negatively impacting patient recovery.

Read More: http://informahealthcare.com/doi/abs/10.3109/09638288.2011.631681