PET scan is widely used nowadays to identify the brain pathological changes during AD (Snyder, Carrillo et al. 2014). CSF biomarker analyses, on the other hand, is quite established for AD diagnostics (Anoop, Singh et al. 2010). These methods, however, have restrictions from been the preferable first- line diagnosis of AD. These restrictions could be countered by the use of blood- based biomarkers. Blood-based diagnosis is less invasive than CSF analysis and less expensive than PET at the same time. Furthermore, clinical routines in the whole world are quite familiar with blood testing (Hampel, O'Bryant et al. 2018). Time- and cost-efficient are not the only advantages of blood-based detection of AD, but more patient’s cooperation is also expected. This means that patients show more follow up commitment in case of blood sample donation compared to e.g. CSF sampling. There is high optimistic that the peripheral blood changes can be captures at early stages of AD pathology. This is because of the increased reports that, at the CNS level, the pathophysiology of AD starts many years before the appearance of clinical signs (Kenny, Jimenez-Mateos et al. 2018). The early detection of AD is very important for therapeutic development. It helps a lot for more effective therapeutic while at the late stages of the disease, irreversible structural changes are happened to the brain and therapeutic interventions are usually too late (Ghezzi, Scarpini et al. 2013, Bagnoli, Piaceri et al. 2014, Marcus, Mena et al. 2014, Kenny, Jimenez-Mateos et al. 2018). Despite these advantages and increased research efforts, the field has been hampered by lack of reproducibility and an unclear strategy to move the basic discovery toward clinical application (O'Bryant, Mielke et al. 2017). One of the biggest issues for developing AD blood-based detection is the phenotypic variability. The variability comes from non-clearly defined disease stages and patient’s variation as well as elderly comorbidities (Hampel, Toschi et al. 2018). An example for this phenotypic variability is the heterogeneity of immune status. Brodin and colleagues conducted a systems-level analysis of 210 healthy twins between 8 and 82 years of age. They showed that there is an extensive heterogeneity in immune-based parameters among healthy individuals (Brodin, Jojic et al. 2015). The high variability for both healthy and patient individuals cannot be handled with single target approaches. This is in our opinion is the main reason that these single target approaches are failing so far to develop blood detection of AD. alternatively, we developed a multi-parametric assay that produced multiple readouts which in turn combined with artificial intelligent algorithm in order to produce predictive model for AD diagnosis. The current study focus mainly on the human immune system which is considerably variable between individuals but relatively stable, over time, for each single individual (Brodin and Davis 2017). We developed high-throughput assay enables simultaneous meas¬urements of cell types, TLR4 activation, inflammasome activation, cytokines release, and RNA expression level from the same blood sample. Such divers’ readouts offer an opportunity for analysing human immune system variation at a global scale, taking a complex machine learning analysis into account. 5.2. Comprehensive assay design to study the immune system In the current study, the immune system was studied via different aspects like cell types, TLR4 activation, inflammasome activation, cytokines release, and RNA expression level. The novelty of the assay comes from testing the blood sample under several conditions including the triggering of TLR pathway and inflammasome machinery or both of them. The assay included the sort of naïve untreated conditions as well. This assay design provided the opportunity to captured any altered phenotype that indicating pathological condition. For example, treating PBMCs with LPS expecting to induce TNF expression. However, capturing TNF expression in untreated PBMCs indicates that within the innate immune system, TLR4 pathway was already activated. The immune systems-level analysis is usually conducted under the steady-state condition (Brodin and Davis 2017, Kaczorowski, Shekhar et al. 2017, Tanamati, Stafuzza et al. 2019). Nevertheless, the multi conditions assay provides much more information than the steady-state measuring assay by revealing a diversity of immune response among individuals. Characteristically, IL-1 release was used as indicator for inflammasome activation (Kesavardhana and Kanneganti 2017, Latz and Duewell 2018, Shao, Cao et al. 2018). In current study, inflammasome activation was monitored during different entry levels. First, inflammasome components are expressed. in process called priming precedes. Then, these components are assembled into multi-protein complex (inflammasome). Finally, inflammasome assembling result into maturation and release Il-1 and eventually lead to a unique cell death called pyroptosis (Man and Kanneganti 2015, Broz and Dixit 2016). During current study, TNFwas monitored as indicators successful priming or TLR activation by the effect of LPS. While aggregations of the inflammasome adapter protein PYCARD were monitored as indicator of inflammasome assembly. Finally, IL-1 release was measured as consequent event of inflammasome activation. Furthermore, the pyroptotic cell death was captured via live tracking of inflammasome activation where the shedding of cell membrane marker (CD14) was captured.